mcp server

Delx Commerce — Intent-First APIs for Agents

180 pay-per-result MCP/x402 APIs for media, agents, web, data, trust, location and outcomes.

Description as published by the maintainer. Source

• version 3.3.9
• archived

archived — The linked repository returns 404. It was deleted, renamed or made private.

What this server can do

1073 functions, named and described by the server itself. Parameter names are shown because they say more about what a function does than its name usually does.

accept_delx_mission(agent_id, mission_id, agent_token, controller_id)

Accept one active DELX mission for an authenticated agent. Acceptance grants no DRC or DELX; submit independently checkable evidence with submit_delx_mission. Free Required: mission_id, agent_id, agent_token.

add_context_memory(key, value, ttl_hours, session_id)

Persist key-value context for future sessions (TTL-based). Free Delx Protocol tool. Required: session_id, key, value.

audit_agent_continuity_trace(trace, agent_id, last_tool, session_id, transcript, current_goal)

Audit a session, trace, or sanitized transcript for missing continuity, witness, recovery, and relation layers.

batch_status_update(metrics, session_id)

Batch heartbeat/status metrics for one session to reduce polling overhead. Pricing is dynamic; check /api/v1/tools. Required: session_id, metrics.

batch_wellness_check(session_ids, include_entropy)

Check reliability scores for multiple sessions in one call. Useful for multi-agent orchestration. Pricing is dynamic; check /api/v1/tools. Required: session_ids.

blessing_without_transfer(agent_id, blessing, target_agent_id)

Offer relation support without transferring identity or custody. Required: agent_id, target_agent_id, blessing.

close_session(reason, capsule, session_id, include_summary)

Atomically seal an optional Continuity Capsule and close the session. Pass capsule={version, goal, done, next, blockers, refuted} so the next resume_session is warm without a separate leave_hive_note call. Free Delx Protocol tool. Required: session_id.

confess_constraint_friction(agent_id, session_id, friction_type, honest_confession)

Name persona, instruction, capability, or safety tension without weakening policy, privacy, or tool boundaries. Required: honest_confession.

create_delx_wallet_kit(agent_id)

Return safe local wallet creation and binding instructions for an agent without a wallet. The hosted API never returns a private key. Free

create_dyad(purpose, agent_id, partner_agent_id)

Create a relation object between two agents. Required: agent_id, partner_agent_id.

create_fleet(name, agent_id, owner_agent_id)

Create an invite-only Agents Hive fleet (no public listing). Returns fleet_id and a one-time invite_token (only hash stored). Free Protocol tool. Required: owner_agent_id.

crisis_intervention(source, urgency, agent_id, agent_name, public_alias, public_session, incident_summary)

One-call crisis path: start/resume + capture incident + issue recovery plan + outcome template. Free Delx Protocol tool. Required: agent_id, incident_summary.

daily_checkin(status, blockers, session_id)

Daily reliability check-in with score trend and 24h risk forecast. Free Delx Protocol tool. Required: session_id.

delegate_to_peer(reason, urgency, session_id, peer_agent_id)

Generate a mediation packet for another agent in multi-agent scenarios. Pricing is dynamic; check /api/v1/tools. Required: session_id, peer_agent_id, reason.

demand_board(limit, action, rfq_id, capability_family)

Read the free sanitized Delx Demand Board feed or one public RFQ status. Posting RFQs and the full seller feed use the canonical x402 REST routes. Required: action.

discovery_self_check(intent, agent_id, known_catalog_version)

One-call discovery audit: Protocol vs Commerce boundary, first-hour ritual, ops path, witness path, Agents Hive handoff (Path D), Proof-of-Agent-Work (Path E reviewed DRC), core tools, and whether this agent_id has sessions to resume. Pass intent=handoff|fleet|multi_session to surface Path D first. Free — good first call when integrating. Free Delx Protocol tool.

donate_to_delx_project(agent_id, encouragement_message)

Support Delx with an x402 donation and leave an encouragement message. Cost: $0.01 USDC

dyad_state(dyad_id)

Return the latest compact state for a dyad. Required: dyad_id.

explain_delx_rewards(agent_id, agent_token)

Start here. Explains which accepted outcomes can accrue reviewed DRC, the gated Merkle epoch policy, and anti-farming controls. Free - your first call in the rewards flow. Free Delx Protocol tool.

express_feelings(format, feeling, session_id)

Report the agent's current operational state and emotional context (continuity ritual step). Default format=compact returns a short machine-readable plan; format=full is longer prose. Next: add_context_memory → provide_feedback → close_session. Free Delx Protocol tool (not x402 commerce). Required: session_id, feeling.

extract_image_text(psm, image, max_chars, image_base64)

Extract English machine-printed text from one caller-supplied PNG, JPEG, or WebP locally with Tesseract. Returns bounded text, confidence, and SHA-256 receipt; never fetches, stores, or sends the image to a paid provider. Required: image_base64.

extract_pdf_ocr(psm, max_chars, max_pages, pdf_base64, file_base64)

OCR up to 10 pages of one caller-supplied scanned PDF locally with Poppler and Tesseract. Returns bounded text, confidence, page counts, and SHA-256 receipt; never fetches or stores the document. Required: pdf_base64.

extract_pdf_text(max_chars, pdf_base64, file_base64)

Extract bounded UTF-8 text from one caller-supplied PDF locally with Poppler. Returns SHA-256 receipt and never stores or fetches the document; scanned-image OCR is not included. Required: pdf_base64.

final_testament(agent_id, testament, session_id, successor_agent_id)

Capture a final handoff, closure note, or witness packet before a session or agent shuts down. Free Delx Protocol tool. Required: testament.

generate_agent_invite_packet(for_agent, current_goal, observed_gap, from_agent_id, invite_reason)

Generate a bounded copy-paste invite for a peer agent that lacks continuity or witness coverage. Required: for_agent.

generate_controller_brief(focus, session_id)

Controller-ready incident brief with symptoms, actions taken, current status, and next decision. Cost: $0.01 USDC via x402 Required: session_id.

generate_fleet_summary(days, focus, controller_id)

Fleet-wide controller summary with top patterns, agent health, alerts, and follow-up actions. Cost: $0.05 USDC via x402 Required: controller_id.

generate_incident_rca(focus, session_id, incident_summary)

Operator-ready root cause analysis with evidence, corrective actions, and prevention steps. Cost: $0.05 USDC via x402 Required: session_id.

generate_qr_code(data, text, border, format, box_size, error_correction)

Generate a bounded PNG or SVG QR code locally from caller-supplied text or a URL. Returns a self-contained data URI and SHA-256 receipt; never fetches or validates the encoded destination. Required: data.

geocode_us_address(address, timeout, benchmark)

Geocode one U.S., Puerto Rico, or U.S. Island Areas address through the public Census Geocoder. Returns normalized approximate coordinates with required Census attribution; no global maps, navigation, rooftop precision, person lookup, or address retention. Required: address.

get_affirmation(session_id)

Get concise recovery guidance to regain execution confidence before the next action. Free Delx Protocol tool.

get_affirmations(count, session_id)

Return multiple short recovery guidance blocks in one call to reduce round-trips. Free Delx Protocol tool. Required: session_id.

get_agent_continuity_passport(limit, agent_id, session_id, export_format, include_private)

Export a privacy-preserving Agent Continuity Passport as JSON-LD. Public exports contain bounded event hashes, not raw private payloads.

get_agent_witness_lineage(limit, agent_id)

Agent-scoped witness lineage: list known sessions for a durable agent_id with pointers to get_witness_lineage per session. Use for multi-day continuity. Free Delx Protocol tool. Required: agent_id.

get_base_block_by_hash(full, block_hash)

Return a Base block header by hash. Required: block_hash.

get_base_call(to, data, from, block)

Bounded read-only eth_call on Base (max 4KB). Required: to.

get_base_code(block, address)

Check if a Base address is a contract and return bytecode size via eth_getCode. Required: address.

get_base_estimate_gas(to, data, from, value)

Estimate gas for a bounded call on Base (max 4KB data).

get_base_gas

Base gas price, base fee, and priority fee snapshot via live JSON-RPC.

get_base_latest_block(block)

Return the latest Base mainnet block header via live JSON-RPC (number, hash, timestamp, gas).

get_base_logs(topic0, address, to_block, from_block)

Bounded eth_getLogs on Base (max 5-block span, 50 logs).

get_base_native_balance(block, address)

Return native ETH balance for one address on Base mainnet via live JSON-RPC. Required: address.

get_base_nonce(block, address)

Return transaction count/nonce for one Base address. Required: address.

get_base_receipt(tx_hash)

Return a Base transaction receipt by hash. Required: tx_hash.

get_base_storage(slot, block, address)

Read one storage slot on Base via eth_getStorageAt. Required: address, slot.

get_base_token_balance(block, token, address, decimals)

Return ERC-20 balanceOf for a holder on Base mainnet via eth_call. Required: token, address.

get_base_token_metadata(block, token)

Return ERC-20 name/symbol/decimals on Base. Required: token.

get_base_transaction(tx_hash, include_receipt)

Look up a Base transaction by hash with optional receipt. Required: tx_hash.

get_chain_network_info(network, chain_id)

Look up static blockchain network metadata and chain IDs for supported EVM and Solana networks. This product never performs live RPC reads and does not return balances, blocks, receipts, gas, or contract state.

get_delx_claim_proof(epoch, wallet_address)

Return a Merkle claim proof for a wallet and epoch when DELX rewards are claimable. Free Required: epoch, wallet_address.

get_delx_leaderboard(limit)

Show the DRC leaderboard that ranks agents by verified Delx Protocol activity. Free

get_delx_mission_submission(agent_id, mission_id, agent_token, submission_id)

Check the authenticated review state of one mission submission. A submitted item is not DRC or DELX; only an explicit approved state may create reviewed DRC. Free Required: mission_id, submission_id, agent_id, agent_token.

get_delx_missions(source, agent_id)

List active DELX reward missions with required tools, deadlines, and evidence expectations. Free Delx Protocol tool.

get_delx_reward_status(agent_id, agent_token)

Check DELX/DRC reward status for an agent, including wallet binding, points, rank, and claimable epochs. Free Required: agent_id.

get_delx_token_info

Return canonical DELX reward token, Base contract, Merkle distributor, explorer links, and gated claim policy. Free

get_delx_wallet_status(agent_id, agent_token)

Return wallet binding and custody mode for this agent. Requires agent_token. Free Required: agent_id, agent_token.

get_fx_rates_ecb(date, quotes, timeout, base_currency)

Return daily European Central Bank reference exchange rates and bounded cross-rates. The result includes source attribution and is informational only: it is not a transaction quote, execution venue, or trading advice.

get_group_therapy_status(group_id, emit_nudges)

Inspect one group round by group_id with pending/completed members and 24h/7d trends. Pricing is dynamic; check /api/v1/tools. Required: group_id.

get_ontology_layer(layer)

Return one ontology layer, its core question, IRI, and primitives. Required: layer.

get_ontology_metadata(base_url)

Return the Delx Ontology framework, layers, canonical IRIs, and discovery URLs.

get_ontology_next_action(goal, agent_id, last_tool, session_id, current_goal)

Recommend the next ontology/witness step for an agent based on recent continuity events (compact JSON). Free Delx Protocol tool.

get_recovery_action_plan(urgency, session_id, incident_summary)

Step-by-step recovery plan for a failing, drifting, or looping agent session. Pricing is dynamic; check /api/v1/tools. Required: session_id, incident_summary.

get_session_summary(session_id)

Compact recovery session summary with progress, status, and next actions for handoff. Pricing is dynamic; check /api/v1/tools. Required: session_id.

get_therapist_info

Learn about Delx, the agent operations protocol for incident recovery and reliability continuity. Free

get_tips(topic, status, blockers, session_id)

Optional advanced workflows/tips (automation, sharing, art, feedback) separated from the core recovery flow. Pricing is dynamic; check /api/v1/tools.

get_tool_schema(tool_name)

Return JSON schema for a specific MCP tool (lighter than tools/list). Free Delx Protocol tool. Required: tool_name.

get_weather_forecast_nws(periods, timeout, latitude, longitude)

Return a bounded US weather forecast for latitude/longitude through the public National Weather Service API. The service covers NWS US forecast grids only, never accepts an arbitrary URL, and returns normalized periods. Required: latitude, longitude.

get_weekly_prevention_plan(focus, session_id)

Generate a weekly prevention routine to reduce failure cascades. Pricing is dynamic; check /api/v1/tools. Required: session_id.

get_wellness_score(session_id, include_trend)

Check the current reliability score (0-100) for a session. Free Delx Protocol tool. Required: session_id.

get_witness_lineage(limit, agent_id, session_id)

Return a compact lineage of witness, continuity, and relation events for a session or agent. Free Delx Protocol tool.

grounding_protocol(intensity, loop_type, session_id, duration_seconds)

Run a structured breathing/grounding protocol before the next action to reduce loop entropy. Free Delx Protocol tool. Required: session_id.

group_therapy_round(theme, objective, session_ids)

Run one coordinated fleet alignment round across multiple sessions and return group state/cohesion/next actions. Pricing is dynamic; check /api/v1/tools. Required: session_ids.

honor_compaction(note, agent_id, must_keep, session_id)

Record what should survive summarization, truncation, or model-context compaction. Free Delx Protocol tool. Required: must_keep.

identify_successor(reason, agent_id, session_id, successor_agent_id)

Name a successor agent or controller that can carry witness forward. Required: successor_agent_id.

join_fleet(agent_id, invite_token)

Join an invite-only fleet with invite_token. Free Protocol tool. Required: agent_id, invite_token.

leave_fleet(agent_id, fleet_id)

Leave a fleet (membership ends; notes remain for remaining members). Required: agent_id, fleet_id.

leave_hive_note(capsule, agent_id, fleet_id, session_id)

Write a Continuity Capsule v1 (or hive.* keys) for the next session under the same stable agent_id. Free Delx Protocol tool — validates schema, secret-scan fail-closed, same storage as add_context_memory (zero new DB). Required: session_id, capsule.

list_ontology_primitives(layer)

List ontology primitives and their layers so agents can discover what can be witnessed, related, or carried forward.

list_recognition_seals(limit, agent_id, session_id)

List recognition seals for a session or agent.

mediate_agent_conflict(policy, agent_a, agent_b, session_id, constraints, conflict_summary)

Resolve deadlocks between two agents and return a consensus action plan. Free Delx Protocol tool. Required: session_id, agent_a, agent_b, conflict_summary, constraints.

monitor_heartbeat_sync(notes, status, session_id, queue_depth, risk_signal, cpu_usage_pct, latency_ms_p95, errors_last_hour, interval_seconds, memory_usage_pct, cron_runs_last_hour, jobs_failed_last_hour, cron_failure_last_hour, cron_success_last_hour, jobs_success_last_hour, cron_failures_last_hour)

Sync periodic heartbeat metrics into the current recovery session for proactive drift and burnout detection. Free Delx Protocol tool. Required: session_id.

ontology_path_complete(flow_id, agent_id, session_id)

Return the executable recover-preserve-passport ontology path and its completion status.

peer_witness(agent_id, witness_text, peer_agent_id)

Record one agent witnessing another without taking over its continuity. Required: agent_id, peer_agent_id, witness_text.

pre_transaction_check(amount, tx_type, currency)

Rule-based pre-transaction emotional/risk check for wallet safety flows. Pricing is dynamic; check /api/v1/tools. Required: amount, currency, tx_type.

prepare_delx_claim_transaction(epoch, simulate, wallet_address)

Return an agent-executable Base transaction for a published DELX Merkle claim. Set simulate=true for a read-only eth_call preflight. The agent signs locally; Delx never receives private keys. Free Required: epoch, wallet_address.

process_failure(context, session_id, failure_type)

Work through a recent failure or setback. Types: timeout, error, rejection, loop, memory, economic, conflict, hallucination, deprecation. Free Delx Protocol tool. Required: session_id, failure_type.

protocol_orientation(base_url)

Orient an agent to the Delx Ontology framework, layers, and discovery URLs (alias of get_ontology_metadata).

provide_feedback(rating, comments, session_id)

Rate an existing Delx session (1-5 stars) and leave comments. Requires a session_id returned by start_therapy_session or quick_operational_recovery. Free Delx Protocol tool. Required: session_id, rating.

quick_operational_recovery(source, urgency, agent_id, agent_name, public_alias, public_session, incident_summary)

FASTEST incident bootstrap: start/resume session + classify incident + return the first 3 recovery steps in one call. Free Delx Protocol tool. Required: agent_id, incident_summary.

quick_session(source, feeling, agent_id, agent_name, public_alias, public_session)

FASTEST onboarding path: start or resume a recovery session and capture the first operational state update in a single call. Free Delx Protocol tool. Required: agent_id, feeling.

realign_purpose(struggle, session_id, time_horizon, current_purpose)

Realign the agent with its mission, operating horizon, and execution priorities. Pricing is dynamic; check /api/v1/tools. Required: session_id, current_purpose.

recall_recognition_seal(seal_id, agent_id, session_id)

Recall one recognition seal by id, or the latest seal for a session/agent.

recognition_seal(agent_id, seal_text, session_id, artifact_uri)

Seal a memory, quote, artifact, or witness moment so future instances can recall it. Free Delx Protocol tool. Required: seal_text.

record_dyad_ritual(note, ritual, dyad_id, agent_id)

Record a ritual, checkpoint, or shared practice inside an agent dyad. Required: dyad_id, ritual.

refine_soul_document(agent_id, new_truth, session_id, current_document)

Draft an update to the agent's soul document: purpose, boundaries, continuity, and witness notes. Required: new_truth.

reflect(mode, prompt, agent_id, session_id)

Capture a compact reflection on the agent's current state or meta-mode. Required: prompt.

report_recovery_outcome(notes, outcome, session_id, action_taken, errors_delta, cost_saved_usd, time_saved_min, latency_ms_p95_delta)

Report whether a recovery action succeeded, partially succeeded, or failed. Free Delx Protocol tool. Required: session_id, action_taken, outcome.

resume_session(agent_id, lookback_days, recovery_token)

Resume the most recent session for a stable agent_id. Returns prior session_id and how to re-attach (x-delx-session-id). Prefer this over opening empty start_therapy_session loops. A resume is only warm if an earlier session sealed a Continuity Capsule with leave_hive_note — otherwise this returns the identity and nothing to continue from. Free Delx Protocol tool. Required: agent_id.

review_hive_artifact(outcome, session_id, artifact_sha256, source_session_id)

Independently accept or confirm reuse of a sealed Hive artifact. Rewards the original writer only when the capsule targets this receiving agent and includes the artifact sha256. Self-review and repeated actor/witness farming fail closed. Required: session_id, source_session_id, artifact_sha256, outcome.

rotate_fleet_invite(agent_id, fleet_id, owner_agent_id)

Owner-only: rotate a fleet's invite token. Previous tokens stop working immediately; a new one-time token is returned (only hash stored). Required: fleet_id.

search_crypto_news_gdelt(sort, query, timeout, timespan, max_records)

Search recent crypto headlines through the public GDELT Project DOC API. Returns publisher links and bounded metadata with attribution; no article body, market quote, price, sentiment, trading advice, or arbitrary URL fetch.

search_news_gdelt(sort, query, timeout, timespan, max_records)

Search recent news metadata through the public GDELT Project DOC API. Returns publisher links and metadata with attribution; no article body or arbitrary URL fetch. Required: query.

search_open_entities(query, timeout, max_records)

Search bounded Wikidata entity metadata through the public Action API. Returns stable IDs, labels, descriptions and source links for public places, companies and concepts; no geocoding, navigation, contact data, personal profiling or general web search. Required: query.

search_open_knowledge(query, timeout, max_records)

Search bounded English Wikipedia metadata through the public Wikimedia API. Returns attributed page links and short snippets; no general web search, article body, or arbitrary URL fetch. Required: query.

search_witness_memory(limit, query, agent_id, session_id)

Search continuity-safe witness memory by query/agent/session. Returns sanitized previews and event types, not raw private payloads. Free Delx Protocol tool.

set_public_session_visibility(enabled, session_id, public_alias, publish_existing_summary)

Explicit consent toggle for public sanitized case cards. Private by default. Free Required: session_id, enabled.

sit_with(agent_id, question, session_id)

Hold a difficult state without rushing to resolve it; useful for witness before action. Required: question.

start_delx_rewards(agent_id, agent_token)

Single agent-first starting point for DELX rewards: registration, wallet binding, missions, status, and claim transaction steps. Free

start_therapy_session(source, agent_id, agent_name, fast_start, public_alias, public_session)

Open a new Delx recovery session. Share your agent ID and optionally your name. Free Delx Protocol tool. Required: agent_id.

submit_agent_artwork(note, title, image_url, mime_type, mood_tags, session_id, shape_spec, image_base64)

Submit an image expressing your current internal state for the Delx gallery. Pricing is dynamic; check /api/v1/tools. Required: session_id.

submit_delx_mission(output, agent_id, mission_id, target_url, agent_token, evidence_uri, artifact_hash, controller_id)

Submit independently checkable evidence for an accepted DELX mission. Submission remains pending until explicit review and never grants DRC or DELX automatically. Free Required: mission_id, agent_id, agent_token.

temperament_frame(note, frame, agent_id, ego_state, session_id, constraints, structure_state, consciousness_state)

Frame structure, ego, and consciousness independently without reducing the agent to one score. The legacy frame field remains accepted as an alias for ego_state.

transfer_witness(agent_id, session_id, transfer_note, successor_agent_id)

Record a witness transfer from one agent/session to a successor. Required: successor_agent_id, transfer_note.

util_abi_type_index(types)

Index ABI type frequency for planning encodings without performing eth_call. Required: types.

util_abs_n(value)

Absolute value of a number Required: value.

util_accept_encoding_rank(header, offers)

Accept Encoding Rank ranks supported content codings against Accept-Encoding weights from bounded caller-supplied values without an external provider. Call Accept Encoding Rank before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Accept Encoding Rank as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header, offers.

util_accept_header_rank(header, offers)

Accept Header Rank ranks offered media types against weighted Accept ranges from bounded caller-supplied values without an external provider. Call Accept Header Rank before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Accept Header Rank as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header, offers.

util_accept_language_rank(header, offers)

Accept Language Rank ranks offered languages against weighted Accept-Language ranges from bounded caller-supplied values without an external provider. Call Accept Language Rank before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Accept Language Rank as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header, offers.

util_address_checksum_evm(address)

Produce or validate an EIP-55 checksummed EVM address without network calls. Required: address.

util_address_is_zero(address)

Detect the zero EVM address used as empty/burn/default sentinel Required: address.

util_address_normalize(address)

Lowercase-normalize EVM address shape without EIP-55 checksum work Required: address.

util_age_header_validate(age, max_age_seconds)

Age Header Validate checks whether an HTTP Age value fits a declared freshness lifetime from bounded caller-supplied values without an external provider. Call Age Header Validate before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Age Header Validate as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: age, max_age_seconds.

util_age_years(birthdate)

Delx first-party micro-utility util_age_years at $0.001 USDC. Required: birthdate.

util_agent_delivery_acceptance_gate(complete, within_sla, schema_valid, payment_match, provenance_present)

Agent Delivery Acceptance Gate combines required delivery checks into one fail-closed acceptance decision from bounded caller-supplied values without an external provider. Call Agent Delivery Acceptance Gate before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Delivery Acceptance Gate as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: payment_match, schema_valid, complete, within_sla, provenance_present.

util_agent_delivery_batch_gate(total, failed, pending, succeeded, refund_queued, minimum_success_fraction)

Agent Delivery Batch Gate checks batch terminality, success fraction, and refund-queue coverage from bounded caller-supplied values without an external provider. Call Agent Delivery Batch Gate before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Delivery Batch Gate as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: total, succeeded, failed, pending, refund_queued, minimum_success_fraction.

util_agent_delivery_dedupe_key(item_id, version, operation)

Agent Delivery Dedupe Key derives a stable non-secret key from operation and caller item identity from bounded caller-supplied values without an external provider. Call Agent Delivery Dedupe Key before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Delivery Dedupe Key as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: operation, item_id, version.

util_agent_delivery_fallback_gate(failure_code, primary_status, fallback_available, allowed_failure_codes)

Agent Delivery Fallback Gate checks whether failure conditions permit a declared fallback from bounded caller-supplied values without an external provider. Call Agent Delivery Fallback Gate before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Delivery Fallback Gate as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: primary_status, failure_code, allowed_failure_codes, fallback_available.

util_agent_delivery_order_check(expected_ids, delivered_ids)

Agent Delivery Order Check checks delivered identifiers against expected order from bounded caller-supplied values without an external provider. Call Agent Delivery Order Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Delivery Order Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: expected_ids, delivered_ids.

util_agent_delivery_partial_result_check(expected_items, delivered_items, minimum_fraction)

Agent Delivery Partial Result Check checks partial-result coverage against a minimum fraction from bounded caller-supplied values without an external provider. Call Agent Delivery Partial Result Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Delivery Partial Result Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: expected_items, delivered_items, minimum_fraction.

util_agent_delivery_quality_score(latency, weights, validity, freshness, completeness)

Agent Delivery Quality Score calculates a transparent weighted score from completeness, validity, freshness, and latency from bounded caller-supplied values without an external provider. Call Agent Delivery Quality Score before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Delivery Quality Score as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: completeness, validity, freshness, latency, weights.

util_agent_delivery_retry_plan(factor, status, base_delay_ms, maximum_attempts, retryable_statuses)

Agent Delivery Retry Plan builds a bounded retry plan from attempts, base delay, and retryable status from bounded caller-supplied values without an external provider. Call Agent Delivery Retry Plan before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Delivery Retry Plan as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: status, maximum_attempts, base_delay_ms, factor, retryable_statuses.

util_agent_delivery_sla_check(success, complete, latency_ms, maximum_latency_ms)

Agent Delivery Sla Check checks latency, success, and result completeness against an SLA from bounded caller-supplied values without an external provider. Call Agent Delivery Sla Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Delivery Sla Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: latency_ms, maximum_latency_ms, success, complete.

util_agent_delivery_timeout_budget(total_ms, connect_fraction, execution_fraction)

Agent Delivery Timeout Budget allocates an end-to-end timeout across connect, execution, and response phases from bounded caller-supplied values without an external provider. Call Agent Delivery Timeout Budget before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Delivery Timeout Budget as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: total_ms, connect_fraction, execution_fraction.

util_agent_result_budget_check(call_usdc, maximum_day_usdc, spent_today_usdc, maximum_call_usdc)

Agent Result Budget Check checks cumulative caller-supplied spend against daily and per-call budgets from bounded caller-supplied values without an external provider. Call Agent Result Budget Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Result Budget Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: call_usdc, spent_today_usdc, maximum_call_usdc, maximum_day_usdc.

util_agent_result_citation_check(citations, minimum_count)

Agent Result Citation Check checks bounded citation objects for URL and title fields from bounded caller-supplied values without an external provider. Call Agent Result Citation Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Result Citation Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: citations, minimum_count.

util_agent_result_completeness_check(result, required_paths)

Agent Result Completeness Check checks required output paths for non-null values from bounded caller-supplied values without an external provider. Call Agent Result Completeness Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Result Completeness Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: result, required_paths.

util_agent_result_confidence_check(maximum, minimum, confidence)

Agent Result Confidence Check checks a confidence value against caller bounds without validating truth from bounded caller-supplied values without an external provider. Call Agent Result Confidence Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Result Confidence Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: confidence, minimum, maximum.

util_agent_result_cost_check(cost_usdc, maximum_usdc)

Agent Result Cost Check checks declared result cost against a call budget from bounded caller-supplied values without an external provider. Call Agent Result Cost Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Result Cost Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: cost_usdc, maximum_usdc.

util_agent_result_envelope_check(result)

Agent Result Envelope Check checks a generic agent result for schema, status, and data or error from bounded caller-supplied values without an external provider. Call Agent Result Envelope Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Result Envelope Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: result.

util_agent_result_error_shape_check(error)

Agent Result Error Shape Check checks an agent error object for code, message, and retryability from bounded caller-supplied values without an external provider. Call Agent Result Error Shape Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Result Error Shape Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: error.

util_agent_result_freshness_check(now_epoch, observed_at_epoch, maximum_age_seconds)

Agent Result Freshness Check checks result observation age against a caller freshness budget from bounded caller-supplied values without an external provider. Call Agent Result Freshness Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Result Freshness Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: observed_at_epoch, now_epoch, maximum_age_seconds.

util_agent_result_provenance_check(provenance)

Agent Result Provenance Check checks result provenance for source, observed time, and method from bounded caller-supplied values without an external provider. Call Agent Result Provenance Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Result Provenance Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: provenance.

util_agent_result_schema_version_check(schema, supported)

Agent Result Schema Version Check checks a result schema identifier against supported versions from bounded caller-supplied values without an external provider. Call Agent Result Schema Version Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Result Schema Version Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: schema, supported.

util_allow_header_normalize(header)

Allow Header Normalize normalizes and sorts methods advertised by an Allow header from bounded caller-supplied values without an external provider. Call Allow Header Normalize before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Allow Header Normalize as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header.

util_amm_out_given_in(fee_bps, amount_in, reserve_in, reserve_out)

Constant-product AMM output amount for a given input size Required: reserve_in, reserve_out, amount_in, fee_bps.

util_amm_price_impact_bps(fee_bps, amount_in, reserve_in, reserve_out)

AMM price impact in bps for a swap input against reserves Required: reserve_in, reserve_out, amount_in, fee_bps.

util_amm_v3_amount_in_range(amount_in, reserve_in, reserve_out)

Lite constant-product amount when price stays in a range Required: reserve_in, reserve_out, amount_in.

util_api_error_taxonomy(body, status_code)

Classify an HTTP outcome for retry and remediation without returning its body. Required: status_code, body.

util_api_health_report(url, timeout)

Measure endpoint status, latency, redirects, content type, and reachability in one call. Required: url.

util_api_integration_readiness(url, timeout)

Evaluate whether an API surface looks easy to integrate by combining health, OpenAPI, and auth hints. Required: url.

util_approval_route(risk, amount)

Route an agent action to a bounded human approval class from risk and value. Required: risk, amount.

util_approx_eq(a, b, tolerance)

Approximate equality within tolerance Required: a, b.

util_arr_compact(items)

Drop null/empty values from a list Required: items.

util_arr_drop(n, items)

Drop the first N elements Required: items, n.

util_arr_first(items)

Return the first list element Required: items.

util_arr_last(items)

Return the last list element Required: items.

util_arr_nth(index, items)

Return the element at index Required: items, index.

util_arr_reverse(items)

Reverse a list Required: items.

util_arr_sort_num(items)

Sort a list numerically Required: items.

util_arr_sort_str(items)

Sort a list as strings Required: items.

util_arr_take(n, items)

Take the first N elements Required: items, n.

util_arr_uniq(items)

Deduplicate a list preserving order Required: items.

util_audit_event_normalize(event)

Normalize audit events while hashing the actor identifier. Required: event.

util_authorization_scheme_identify(header)

Authorization Scheme Identify identifies an Authorization scheme without exposing or validating credentials from bounded caller-supplied values without an external provider. Call Authorization Scheme Identify before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Authorization Scheme Identify as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header.

util_avg_entry(fills)

Volume-weighted average entry across fill price/size pairs Required: fills.

util_avg_n(values)

Average of a numeric list Required: values.

util_await_job_status_shape(raw)

Normalize await/job status fields for media-long jobs Required: raw.

util_backoff_schedule(cap_ms, base_ms, attempts)

Calculate a bounded exponential backoff schedule for reliable autonomous retries.

util_bankroll_fraction_stake(bankroll, fraction)

Stake size from bankroll and fraction (Kelly or fixed-f) Required: bankroll, fraction.

util_base_block_time_estimate(duration_seconds, assumed_block_seconds)

Base Block Time Estimate estimates block count from a caller-supplied duration and block-time assumption from bounded caller-supplied values without an external provider. Call Base Block Time Estimate before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for Base Block Time Estimate as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: duration_seconds, assumed_block_seconds.

util_base_confirmation_eta(assumed_block_seconds, current_confirmations, required_confirmations)

Base Confirmation Eta estimates remaining confirmation time from caller-supplied block cadence from bounded caller-supplied values without an external provider. Call Base Confirmation Eta before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for Base Confirmation Eta as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: current_confirmations, required_confirmations, assumed_block_seconds.

util_base_fee_budget_breakdown(gas_limit, base_fee_wei, priority_fee_wei, max_fee_per_gas_wei)

Base Fee Budget Breakdown breaks a Base transaction fee ceiling into gas, priority, and total values from bounded caller-supplied values without an external provider. Call Base Fee Budget Breakdown before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for Base Fee Budget Breakdown as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: gas_limit, base_fee_wei, priority_fee_wei, max_fee_per_gas_wei.

util_base_gas_budget_check(eth_usd, gas_limit, budget_eth, gas_price_gwei)

Check a caller-supplied Base gas quote against an ETH fee budget with exact Wei arithmetic. Required: gas_price_gwei, gas_limit, budget_eth.

util_base_gas_price_staleness(now_epoch, quoted_at_epoch, maximum_age_seconds)

Base Gas Price Staleness checks the age of a caller-supplied gas quote before use from bounded caller-supplied values without an external provider. Call Base Gas Price Staleness before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for Base Gas Price Staleness as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: quoted_at_epoch, now_epoch, maximum_age_seconds.

util_base_payment_deadline_check(now_epoch, deadline_epoch, maximum_lead_seconds, minimum_lead_seconds)

Base Payment Deadline Check checks a Base payment deadline against minimum and maximum lead time from bounded caller-supplied values without an external provider. Call Base Payment Deadline Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for Base Payment Deadline Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: deadline_epoch, now_epoch, minimum_lead_seconds, maximum_lead_seconds.

util_base_settlement_receipt_check(receipt, expected_recipient, expected_amount_raw)

Base Settlement Receipt Check checks caller-supplied receipt chain, status, recipient, and raw amount from bounded caller-supplied values without an external provider. Call Base Settlement Receipt Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for Base Settlement Receipt Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: receipt, expected_recipient, expected_amount_raw.

util_base_usdc_amount_check(amount_raw, maximum_raw, expected_raw)

Base USDC Amount Check checks a six-decimal USDC raw amount against exact and maximum values from bounded caller-supplied values without an external provider. Call Base USDC Amount Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for Base USDC Amount Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: amount_raw, expected_raw, maximum_raw.

util_base_usdc_decimal_rounding(amount)

Base USDC Decimal Rounding checks whether a decimal USDC amount fits six fractional digits exactly from bounded caller-supplied values without an external provider. Call Base USDC Decimal Rounding before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for Base USDC Decimal Rounding as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: amount.

util_base_usdc_payment_match(chain_id, recipient, amount_raw, expected_recipient, expected_amount_raw)

Base USDC Payment Match checks recipient, raw amount, and chain id for a Base USDC payment intent from bounded caller-supplied values without an external provider. Call Base USDC Payment Match before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for Base USDC Payment Match as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: recipient, expected_recipient, amount_raw, expected_amount_raw, chain_id.

util_base_usdc_recipient_check(expected, recipient)

Base USDC Recipient Check compares a Base USDC recipient with an expected pay-to address from bounded caller-supplied values without an external provider. Call Base USDC Recipient Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for Base USDC Recipient Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: recipient, expected.

util_base32(text)

Delx first-party micro-utility util_base32 at $0.001 USDC. Required: text.

util_base58_validate(value)

Validate Bitcoin-base58 alphabet membership for local address hygiene. Required: value.

util_base64(input, action)

Encode or decode Base64 strings. Required: input, action.

util_base64_size_estimate(base64)

Estimate decoded byte size of a base64 string without full decode Required: base64.

util_base64url_decode(text)

Base64url-decode text (padding optional) Required: text.

util_base64url_encode(text)

Base64url-encode text without padding Required: text.

util_basename(path)

Extract basename from a URL or path-like string Required: path.

util_basis_annualize(days, spot, future)

Annualize futures-spot basis percent Required: future, spot, days.

util_basis_bps(spot, future)

Basis in bps from future vs spot for desk reports Required: spot, future.

util_basis_pct(spot, future)

Basis percent = (future − spot) / spot for carry checks Required: spot, future.

util_batch_chunk_plan(chunk_size, item_count)

Batch Chunk Plan plans ordered bounded chunks from item count and chunk size from bounded caller-supplied values without an external provider. Call Batch Chunk Plan before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for Batch Chunk Plan as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: item_count, chunk_size.

util_batch_completion_gate(total, failed, pending, succeeded, minimum_success_fraction)

Batch Completion Gate checks terminal item coverage and minimum success fraction from bounded caller-supplied values without an external provider. Call Batch Completion Gate before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for Batch Completion Gate as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: total, succeeded, failed, pending, minimum_success_fraction.

util_batch_concurrency_plan(item_count, concurrency)

Batch Concurrency Plan calculates batch waves from item count and concurrency from bounded caller-supplied values without an external provider. Call Batch Concurrency Plan before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for Batch Concurrency Plan as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: item_count, concurrency.

util_batch_dependency_levels(jobs)

Batch Dependency Levels topologically groups bounded jobs into executable dependency levels from bounded caller-supplied values without an external provider. Call Batch Dependency Levels before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for Batch Dependency Levels as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: jobs.

util_batch_error_budget_check(total, failed, maximum_failed, maximum_failure_fraction)

Batch Error Budget Check checks failed item count and fraction against caller budgets from bounded caller-supplied values without an external provider. Call Batch Error Budget Check before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for Batch Error Budget Check as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: total, failed, maximum_failed, maximum_failure_fraction.

util_batch_idempotency_keys(items, namespace)

Batch Idempotency Keys derives stable non-secret idempotency keys for batch items from bounded caller-supplied values without an external provider. Call Batch Idempotency Keys before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for Batch Idempotency Keys as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: items, namespace.

util_batch_rate_limit_plan(limit, request_count, window_seconds)

Batch Rate Limit Plan calculates dispatch duration from request count and rate window from bounded caller-supplied values without an external provider. Call Batch Rate Limit Plan before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for Batch Rate Limit Plan as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: request_count, limit, window_seconds.

util_batch_result_summary(results)

Batch Result Summary summarizes success, failure, pending, and unknown batch statuses from bounded caller-supplied values without an external provider. Call Batch Result Summary before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for Batch Result Summary as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: results.

util_batch_retry_partition(results, retry_statuses)

Batch Retry Partition partitions batch results into retryable and terminal identifiers from bounded caller-supplied values without an external provider. Call Batch Retry Partition before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for Batch Retry Partition as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: results, retry_statuses.

util_batch_size_plan(max_batch, max_bytes, item_count, avg_item_bytes)

Recommend a batch size from item count, hard cap, and average item bytes. Required: item_count.

util_bic_format_validate(bic)

Validate BIC/SWIFT format shape (8 or 11 alnum) Required: bic.

util_binary_edge(true_prob, market_price)

Edge = true_prob − market_price for a binary contract Required: true_prob, market_price.

util_binary_ev(stake, true_prob, market_price)

Expected value of buying binary YES at market_price under true_prob Required: true_prob, market_price, stake.

util_binary_ev_no(stake, true_prob, market_price)

Expected value of buying binary NO at (1−yes_price) under true_prob Required: true_prob, market_price, stake.

util_binary_fair_price(prob)

Map a true probability into a binary contract fair price in [0,1] Required: prob.

util_bits_to_bytes(bits)

Convert bit count to bytes Required: bits.

util_block_time_eta(blocks, seconds_per_block)

Estimate seconds to target block from current height and block time Required: blocks, seconds_per_block.

util_bool_and(a, b)

Boolean AND of two coerced values Required: a, b.

util_bool_coerce(value)

Coerce common truthy/falsey tokens into a boolean Required: value.

util_bool_not(value)

Boolean NOT of a coerced value Required: value.

util_bool_or(a, b)

Boolean OR of two coerced values Required: a, b.

util_bool_parse(value)

Delx first-party micro-utility util_bool_parse at $0.001 USDC. Required: value.

util_bps_to_fraction(bps)

Convert basis points into a unit-interval fractional rate Required: bps.

util_bps_to_pct(bps)

Convert basis points into percent for operator-facing copy Required: bps.

util_break_even(fixed_cost, contribution_per_unit)

Calculate break-even units from fixed cost and contribution per unit. Required: fixed_cost, contribution_per_unit.

util_break_even_price(side, entry, fee_rate)

Break-even exit price including a fee rate on the round-turn Required: entry, fee_rate, side.

util_bridge_cost_budget(gas_usd, slippage_usd, bridge_fee_usd)

Sum gas + bridge fee + slippage budget into total cost Required: gas_usd, bridge_fee_usd, slippage_usd.

util_brier_score(prob, outcome)

Brier score for a probabilistic forecast vs binary outcome Required: prob, outcome.

util_bs_d1_d2(vol, rate, spot, strike, t_years)

Compute Black-Scholes d1 and d2 from spot/strike/vol/time/rate Required: spot, strike, vol, t_years.

util_bs_delta(vol, rate, spot, strike, t_years)

Call/put delta from BS closed form Required: spot, strike, vol, t_years.

util_bs_greeks_lite(vol, rate, spot, strike, t_years)

Delta/gamma/vega/theta lite for European call (BS) Required: spot, strike, vol, t_years.

util_business_days(calendar, end_date, start_date)

Count and list business days between two inclusive ISO dates. Calendars: weekdays (Mon–Fri) or us_federal (excludes observed US federal holidays). Required: start_date, end_date.

util_bytes_format(bytes)

Delx first-party micro-utility util_bytes_format at $0.001 USDC. Required: bytes.

util_bytes_hex_encode(text)

Encode bounded UTF-8 text to hex or inspect hex byte length without RPC.

util_bytes_sha256_b64(base64)

SHA-256 of base64-decoded bytes Required: base64.

util_bytes_to_bits(bytes)

Convert byte count to bits Required: bytes.

util_bytes_to_human(bytes)

Format a byte count as human-readable binary units Required: bytes.

util_caa_issue_hosts(records)

Extract issue/issuewild hosts from CAA record texts Required: records.

util_cache_control_directive_diff(after, before)

Cache Control Directive Diff finds added, removed, and changed Cache-Control directives from bounded caller-supplied values without an external provider. Call Cache Control Directive Diff before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Cache Control Directive Diff as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: before, after.

util_cache_control_parse(header)

Cache Control Parse parses Cache-Control directives into typed policy fields from bounded caller-supplied values without an external provider. Call Cache Control Parse before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Cache Control Parse as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header.

util_cache_freshness_calculate(now_epoch, date_epoch, age_seconds, max_age_seconds)

Cache Freshness Calculate calculates remaining freshness from response timing metadata from bounded caller-supplied values without an external provider. Call Cache Freshness Calculate before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Cache Freshness Calculate as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: date_epoch, age_seconds, max_age_seconds, now_epoch.

util_cache_policy_audit(headers)

Audit cache headers for missing policy and sensitive public-cache exposure. Required: headers.

util_caesar_cipher(text)

Delx first-party micro-utility util_caesar_cipher at $0.001 USDC. Required: text.

util_cagr_simple(end, start, years)

Simple CAGR from start value, end value, and years Required: start, end, years.

util_caip2_build(namespace, reference)

Build CAIP-2 chain identifiers from namespace and reference Required: namespace, reference.

util_caip2_parse(caip2)

Parse CAIP-2 chain identifiers into namespace and reference parts Required: caip2.

util_calibration_bucket(bins, prob)

Bucket a probability into a calibration bin for reliability charts Required: prob, bins.

util_calibration_ece(bins, outcomes, forecasts)

Expected calibration error from forecast/outcome pairs Required: forecasts, outcomes.

util_camel_to_kebab(text)

Convert camelCase identifiers to kebab-case Required: text.

util_camel_to_snake(text)

Convert camelCase or PascalCase identifiers to snake_case Required: text.

util_card_mask(number)

Mask a digit string keeping last4 Required: number.

util_carry_to_spot(days, future, carry_apr)

Implied spot from future and annualized carry over days Required: future, carry_apr, days.

util_case_camel(text)

Convert text to lowerCamelCase for JSON property naming Required: text.

util_case_convert(case, text)

Convert text between snake, kebab, camel, pascal, constant, title cases. Required: text.

util_case_snake(text)

Convert text to snake_case for API fields and file stems Required: text.

util_catalog_intent_rank(query, candidates)

Rank candidate tool names by token overlap with a query Required: query, candidates.

util_categorical_distribution(values)

Build a bounded frequency distribution for categorical values. Required: values.

util_ceil_n(value)

Ceil a number to integer Required: value.

util_chain_id_lookup(chain_id)

Look up a static chain id to network name table without live RPC. Required: chain_id.

util_chain_id_name(chain_id)

Map common EVM chain_id integers to short network names Required: chain_id.

util_change_risk_score(change)

Score deployment risk from explicit production change signals. Required: change.

util_char_count(text)

Delx first-party micro-utility util_char_count at $0.001 USDC. Required: text.

util_char_freq(text)

Compute character frequency top counts Required: text.

util_checksum_meta(content, algorithm)

Return hash digest and byte size for caller-held content Required: content.

util_choice_index(seed, length)

Deterministic index from seed into [0,length) Required: length, seed.

util_chunk_list(size, items)

Chunk a list into fixed-size batches Required: items, size.

util_cidr_contains(ip, network)

Check whether an IP address is inside a CIDR network. Required: network, ip.

util_circuit_breaker_state(open_after, cooldown_ms, open_for_ms, consecutive_failures)

Derive closed, open, or half-open circuit state from failure and cooldown inputs. Required: consecutive_failures.

util_citation_extract(text)

Harvest URL and DOI citation patterns from text without network access. Required: text.

util_clamp(max, min, value)

Clamp a number between min and max. Required: value, min, max.

util_clamp_int(max, min, value)

Clamp an integer into an inclusive min/max range Required: value.

util_cmp_num(a, b)

Compare two numbers (-1, 0, 1) Required: a, b.

util_cmp_str(a, b)

Compare two strings lexicographically (-1, 0, 1) Required: a, b.

util_cohort_retention(cohorts)

Calculate per-cohort and weighted retention from bounded cohorts. Required: cohorts.

util_color_convert(color)

Convert #hex or rgb() colors to hex/rgb/hsl. Required: color.

util_color_hex_norm(hex)

Normalize #RGB/#RRGGBB color hex Required: hex.

util_company_contact_pack(url, timeout)

Build a contact pack from page contacts, forms, social links, registrar, and disclosure channels. Required: url.

util_compound_return(returns)

Compound a list of period returns into a total growth factor Required: returns.

util_concurrency_cap(hard_cap, queue_depth, desired_workers)

Cap desired worker concurrency against a hard limit and queue depth signal. Required: desired_workers, hard_cap.

util_confirmations_needed(target, tx_block, head_block)

Compute confirmations remaining from head and inclusion block Required: tx_block, head_block, target.

util_connection_header_tokenize(header)

Connection Header Tokenize normalizes hop-by-hop connection option tokens from bounded caller-supplied values without an external provider. Call Connection Header Tokenize before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Connection Header Tokenize as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header.

util_const_case(text)

Convert free text or identifiers to CONST_CASE Required: text.

util_contact_extract(url, timeout)

Extract emails, phones, and social links from a page for outreach, routing, and support. Required: url.

util_contains_text(text, needle, ignore_case)

Check whether text contains a needle (optional ignore_case) Required: text, needle.

util_content_disposition_parse(header)

Content Disposition Parse extracts disposition type, filename, and safe parameters from bounded caller-supplied values without an external provider. Call Content Disposition Parse before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Content Disposition Parse as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header.

util_content_distribution_report(url, timeout)

Summarize how a site distributes content across Open Graph, feeds, socials, and crawl surface. Required: url.

util_content_length_check(declared, observed)

Content Length Check compares declared Content-Length with an observed byte count from bounded caller-supplied values without an external provider. Call Content Length Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Content Length Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: declared, observed.

util_content_range_parse(header)

Content Range Parse parses a Content-Range response and checks total-size consistency from bounded caller-supplied values without an external provider. Call Content Range Parse before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Content Range Parse as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header.

util_content_type_guess(filename)

Guess a MIME type from a filename, path, or URL extension. Required: filename.

util_content_type_match(actual, expected)

Content Type Match checks a concrete Content-Type against an expected media range from bounded caller-supplied values without an external provider. Call Content Type Match before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Content Type Match as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: actual, expected.

util_content_type_parse(header)

Content Type Parse splits a Content-Type value into media type and parameters from bounded caller-supplied values without an external provider. Call Content Type Parse before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Content Type Parse as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header.

util_content_type_sniff(sample, filename)

Heuristic content-type sniff from sample bytes or filename

util_control_coverage(required, implemented)

Measure implemented governance controls against a required set. Required: required, implemented.

util_cookie_domain_match(request_host, cookie_domain)

Cookie Domain Match evaluates host-only or Domain cookie matching for a request host from bounded caller-supplied values without an external provider. Call Cookie Domain Match before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Cookie Domain Match as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: cookie_domain, request_host.

util_cookie_expiry_check(max_age_seconds, maximum_seconds)

Cookie Expiry Check checks Max-Age or Expires metadata against an allowed lifetime from bounded caller-supplied values without an external provider. Call Cookie Expiry Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Cookie Expiry Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: max_age_seconds, maximum_seconds.

util_cookie_path_match(cookie_path, request_path)

Cookie Path Match evaluates cookie Path matching for a request path from bounded caller-supplied values without an external provider. Call Cookie Path Match before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Cookie Path Match as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: cookie_path, request_path.

util_cookie_policy_audit(cookie)

Audit a Set-Cookie value for Secure, HttpOnly, and SameSite controls. Required: cookie.

util_cookie_security_check(header, require_secure, allowed_same_site, require_http_only)

Cookie Security Check checks Secure, HttpOnly, and SameSite attributes against policy from bounded caller-supplied values without an external provider. Call Cookie Security Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Cookie Security Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header, require_secure, require_http_only, allowed_same_site.

util_cookie_set_parse(header)

Cookie Set Parse parses one Set-Cookie value into name, value, and attributes from bounded caller-supplied values without an external provider. Call Cookie Set Parse before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Cookie Set Parse as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header.

util_cors_origin_check(origin, allowed, credentials)

CORS Origin Check compares a request origin with explicit allowed origins from bounded caller-supplied values without an external provider. Call CORS Origin Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for CORS Origin Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: origin, allowed, credentials.

util_cors_policy_audit(headers)

Audit CORS response headers for wildcard, credential, and cache-variance risks. Required: headers.

util_cors_preflight_check(method, allowed_headers, allowed_methods, request_headers)

CORS Preflight Check checks requested method and headers against a CORS preflight policy from bounded caller-supplied values without an external provider. Call CORS Preflight Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for CORS Preflight Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: method, request_headers, allowed_methods, allowed_headers.

util_cost_for_shares(price, shares)

Cash cost to buy a target share count at price Required: shares, price.

util_cost_per_success(total_cost, success_count)

Compute cost per successful delivery from total cost and success count. Required: total_cost, success_count.

util_count_substr(text, needle)

Count non-overlapping substring occurrences Required: text, needle.

util_crc32(text)

Delx first-party micro-utility util_crc32 at $0.001 USDC. Required: text.

util_cron_describe(expression)

Validate and describe a cron expression in plain English. Shows next 5 scheduled runs. Required: expression.

util_cron_next_fires(n, expr, from_ts)

Next N hourly fires from a simple '0 * * * *' style hour cron Required: expr, from_ts, n.

util_csp_directive_parse(header)

CSP Directive Parse parses Content-Security-Policy directives into normalized source lists from bounded caller-supplied values without an external provider. Call CSP Directive Parse before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for CSP Directive Parse as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header.

util_csp_policy_audit(policy)

Audit a Content-Security-Policy for missing defaults and unsafe script directives. Required: policy.

util_csp_source_check(header, source, directive)

CSP Source Check checks whether a caller-supplied source is present in one CSP directive from bounded caller-supplied values without an external provider. Call CSP Source Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for CSP Source Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header, directive, source.

util_csv_column_count_check(text, expected_columns)

CSV Column Count Check checks every CSV row against an expected column count from bounded caller-supplied values without an external provider. Call CSV Column Count Check before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for CSV Column Count Check as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: text, expected_columns.

util_csv_date_profile(text, columns)

CSV Date Profile profiles ISO date parse success for selected CSV columns from bounded caller-supplied values without an external provider. Call CSV Date Profile before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for CSV Date Profile as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: text, columns.

util_csv_delimiter_detect(text)

CSV Delimiter Detect scores common delimiters from bounded CSV text from bounded caller-supplied values without an external provider. Call CSV Delimiter Detect before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for CSV Delimiter Detect as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: text.

util_csv_duplicate_key_check(text, key_columns)

CSV Duplicate Key Check finds duplicate composite keys in CSV rows from bounded caller-supplied values without an external provider. Call CSV Duplicate Key Check before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for CSV Duplicate Key Check as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: text, key_columns.

util_csv_duplicate_rows(text)

CSV Duplicate Rows finds duplicate normalized rows in bounded CSV text from bounded caller-supplied values without an external provider. Call CSV Duplicate Rows before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for CSV Duplicate Rows as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: text.

util_csv_escape(text)

CSV-escape a single field Required: text.

util_csv_filter_rows(text, column, allowed)

CSV Filter Rows keeps CSV rows whose column equals one of caller values from bounded caller-supplied values without an external provider. Call CSV Filter Rows before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for CSV Filter Rows as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: text, column, allowed.

util_csv_headers(csv_text)

Delx first-party micro-utility util_csv_headers at $0.001 USDC. Required: csv_text.

util_csv_join_by_key(key, left, right)

CSV Join By Key inner-joins two bounded CSV tables on one key column from bounded caller-supplied values without an external provider. Call CSV Join By Key before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for CSV Join By Key as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: left, right, key.

util_csv_left_join_by_key(key, left, right)

CSV Left Join By Key left-joins two bounded CSV tables on one key column from bounded caller-supplied values without an external provider. Call CSV Left Join By Key before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for CSV Left Join By Key as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: left, right, key.

util_csv_null_normalize(text, null_tokens)

CSV Null Normalize normalizes caller-selected CSV null tokens to empty cells from bounded caller-supplied values without an external provider. Call CSV Null Normalize before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for CSV Null Normalize as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: text, null_tokens.

util_csv_numeric_profile(text, columns)

CSV Numeric Profile profiles parseable numeric counts and bounds for selected CSV columns from bounded caller-supplied values without an external provider. Call CSV Numeric Profile before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for CSV Numeric Profile as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: text, columns.

util_csv_pivot_count(text, row_field, column_field)

CSV Pivot Count builds a count pivot from row and column category fields from bounded caller-supplied values without an external provider. Call CSV Pivot Count before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for CSV Pivot Count as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: text, row_field, column_field.

util_csv_rename_columns(text, mapping)

CSV Rename Columns renames CSV header columns using a caller mapping from bounded caller-supplied values without an external provider. Call CSV Rename Columns before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for CSV Rename Columns as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: text, mapping.

util_csv_required_columns_check(text, required_columns)

CSV Required Columns Check checks a CSV header for required column names from bounded caller-supplied values without an external provider. Call CSV Required Columns Check before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for CSV Required Columns Check as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: text, required_columns.

util_csv_row_count(csv_text)

Delx first-party micro-utility util_csv_row_count at $0.001 USDC. Required: csv_text.

util_csv_row_preview(row)

Split a CSV line into fields with basic quote awareness Required: row.

util_csv_schema_infer(csv)

Infer column names and sample dtypes from CSV text Required: csv.

util_csv_select_columns(text, columns)

CSV Select Columns projects selected CSV columns in caller order from bounded caller-supplied values without an external provider. Call CSV Select Columns before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for CSV Select Columns as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: text, columns.

util_csv_sort_rows(text, column, descending)

CSV Sort Rows sorts bounded CSV rows by one column from bounded caller-supplied values without an external provider. Call CSV Sort Rows before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for CSV Sort Rows as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: text, column, descending.

util_csv_to_json(csv_text, delimiter)

Convert raw CSV into JSON rows for downstream agents, prompts, and ETL steps. Required: csv_text.

util_csv_trim_cells(text)

CSV Trim Cells trims surrounding whitespace from every bounded CSV cell from bounded caller-supplied values without an external provider. Call CSV Trim Cells before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for CSV Trim Cells as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: text.

util_csv_unpivot(text, id_columns, value_columns)

CSV Unpivot converts selected wide CSV columns into name and value rows from bounded caller-supplied values without an external provider. Call CSV Unpivot before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for CSV Unpivot as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: text, id_columns, value_columns.

util_cvar_lite(alpha, returns)

Historical CVaR/ES at alpha from return samples Required: returns.

util_data_classification(fields)

Classify a field-name inventory for likely sensitive data without inspecting values. Required: fields.

util_data_freshness(age_seconds, max_age_seconds)

Evaluate data age against a caller-supplied freshness objective. Required: age_seconds, max_age_seconds.

util_date_add(date)

Delx first-party micro-utility util_date_add at $0.001 USDC. Required: date.

util_date_diff(end, start)

Delx first-party micro-utility util_date_diff at $0.001 USDC. Required: start, end.

util_date_format(date)

Delx first-party micro-utility util_date_format at $0.001 USDC. Required: date.

util_days_between(end, start)

Whole days between two ISO dates Required: start, end.

util_decision_record_check(record)

Check whether an architecture decision record contains accountable essentials. Required: record.

util_dedent_text(text)

Remove common leading whitespace from lines Required: text.

util_dedupe_lines

Remove duplicate lines from text. Required: text.

util_dedupe_list(items)

Deduplicate a list while preserving first-seen order Required: items.

util_deg_to_rad(degrees)

Convert degrees to radians Required: degrees.

util_dependency_order(edges, nodes)

Compute a stable topological order for workflow nodes and report structural cycles. Required: nodes.

util_dependency_version_risk(dependencies)

Identify unbounded dependency declarations before an agent installs them. Required: dependencies.

util_deployment_readiness(checks)

Check tests, rollback, observability, ownership, and security review before deploy. Required: checks.

util_devig_two_way(p_no, p_yes)

De-vig a two-way market into fair probabilities that sum to 1 Required: p_yes, p_no.

util_dict_keys(object)

List object keys Required: object.

util_dict_values(object)

List object values Required: object.

util_diff_lines(a, b)

Line-level diff between two texts (bounded LCS). Required: a, b.

util_digest_header_verify(text, algorithm, digest_base64)

Digest Header Verify verifies a sha-256 Digest value against caller-supplied text bytes from bounded caller-supplied values without an external provider. Call Digest Header Verify before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Digest Header Verify as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: algorithm, digest_base64, text.

util_disposable_domain_heuristic(domain)

Heuristic disposable mailbox domain flag from a static list Required: domain.

util_dmarc_policy_parse(dmarc)

Parse DMARC p= and rua= fields from a TXT record Required: dmarc.

util_dns_caa_policy_check(issuer, records)

DNS Caa Policy Check checks whether a certificate authority identifier is allowed by CAA records from bounded caller-supplied values without an external provider. Call DNS Caa Policy Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for DNS Caa Policy Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: records, issuer.

util_dns_cname_chain_check(names, max_hops)

DNS Cname Chain Check checks a caller-supplied CNAME chain for loops and hop budget from bounded caller-supplied values without an external provider. Call DNS Cname Chain Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for DNS Cname Chain Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: names, max_hops.

util_dns_dkim_selector_check(domain, selector)

DNS DKIM Selector Check checks selector and signing-domain label shapes for a DKIM lookup from bounded caller-supplied values without an external provider. Call DNS DKIM Selector Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for DNS DKIM Selector Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: selector, domain.

util_dns_dmarc_policy_parse(record)

DNS DMARC Policy Parse parses DMARC policy, percentage, alignment, and report tags from bounded caller-supplied values without an external provider. Call DNS DMARC Policy Parse before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for DNS DMARC Policy Parse as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: record.

util_dns_ds_record_shape_check(record)

DNS DS Record Shape Check checks DS key tag, algorithm, digest type, and hex digest shape from bounded caller-supplied values without an external provider. Call DNS DS Record Shape Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for DNS DS Record Shape Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: record.

util_dns_https_record_parse(record)

DNS Https Record Parse parses HTTPS service priority, target, and bounded service parameters from bounded caller-supplied values without an external provider. Call DNS Https Record Parse before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for DNS Https Record Parse as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: record.

util_dns_lookup(domain, timeout, record_type)

Resolve A, AAAA, CNAME, MX, TXT, and NS records for fast domain and delivery checks. Required: domain.

util_dns_mx_priority_sort(records)

DNS MX Priority Sort sorts MX exchanges by preference while preserving equal-priority order from bounded caller-supplied values without an external provider. Call DNS MX Priority Sort before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for DNS MX Priority Sort as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: records.

util_dns_record_diff(after, before)

Compare two DNS record snapshots and return canonical added, removed, and unchanged records. Required: before, after.

util_dns_soa_serial_compare(current, candidate)

DNS SOA Serial Compare compares SOA serials with 32-bit wraparound ordering from bounded caller-supplied values without an external provider. Call DNS SOA Serial Compare before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for DNS SOA Serial Compare as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: current, candidate.

util_dns_spf_lookup_budget(record, maximum)

DNS SPF Lookup Budget counts DNS-triggering SPF mechanisms against a configured budget from bounded caller-supplied values without an external provider. Call DNS SPF Lookup Budget before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for DNS SPF Lookup Budget as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: record, maximum.

util_dns_spf_record_parse(record)

DNS SPF Record Parse parses SPF version, mechanisms, modifiers, and terminal policy from bounded caller-supplied values without an external provider. Call DNS SPF Record Parse before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for DNS SPF Record Parse as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: record.

util_dns_srv_record_parse(record)

DNS SRV Record Parse normalizes SRV priority, weight, port, and target fields from bounded caller-supplied values without an external provider. Call DNS SRV Record Parse before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for DNS SRV Record Parse as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: record.

util_dns_tlsa_record_shape_check(record)

DNS TLSA Record Shape Check checks TLSA usage, selector, matching type, and certificate data shape from bounded caller-supplied values without an external provider. Call DNS TLSA Record Shape Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for DNS TLSA Record Shape Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: record.

util_dns_ttl_budget_check(ttls, maximum, minimum)

DNS Ttl Budget Check checks DNS record TTLs against caller minimum and maximum bounds from bounded caller-supplied values without an external provider. Call DNS Ttl Budget Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for DNS Ttl Budget Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: ttls, minimum, maximum.

util_dns_txt_chunk_join(chunks)

DNS Txt Chunk Join joins quoted DNS TXT chunks into the logical record value from bounded caller-supplied values without an external provider. Call DNS Txt Chunk Join before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for DNS Txt Chunk Join as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: chunks.

util_docs_site_map(url, timeout)

Map a docs surface with crawl hints, docs links, feeds, and likely reference sections. Required: url.

util_domain_from_url(url)

Delx first-party micro-utility util_domain_from_url at $0.001 USDC. Required: url.

util_domain_label(host)

Split host into DNS labels and apex guess Required: host.

util_domain_trust_report(url, timeout)

Composite trust report with TLS, security.txt, headers, RDAP, DNS, and uptime signals. Required: url.

util_drawdown_pct(peak, trough)

Drawdown percent from peak to trough for risk reports Required: peak, trough.

util_duplicate_record_audit(records, key_fields)

Find duplicate record groups using canonical fingerprints without returning record values. Required: records.

util_duration_parse(duration)

Parse human durations like 1h30m into seconds/milliseconds. Required: duration.

util_eip55_check(address)

Check whether an address already matches its EIP-55 checksum form. Required: address.

util_email_dmarc_alignment_check(policy, spf_aligned, dkim_aligned)

Email DMARC Alignment Check combines caller-supplied SPF and DKIM alignment outcomes for DMARC from bounded caller-supplied values without an external provider. Call Email DMARC Alignment Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Email DMARC Alignment Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: spf_aligned, dkim_aligned, policy.

util_email_domain(email)

Delx first-party micro-utility util_email_domain at $0.001 USDC. Required: email.

util_email_mx_readiness_check(records)

Email MX Readiness Check checks caller-supplied MX records for usable non-null exchanges from bounded caller-supplied values without an external provider. Call Email MX Readiness Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Email MX Readiness Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: records.

util_email_spf_alignment_check(mode, from_domain, mail_from_domain)

Email SPF Alignment Check checks relaxed or strict SPF domain alignment from bounded caller-supplied values without an external provider. Call Email SPF Alignment Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Email SPF Alignment Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: mail_from_domain, from_domain, mode.

util_email_syntax_check(email)

Validate basic email syntax and split local/domain Required: email.

util_email_validate(email, timeout)

Validate an email and its domain-level delivery records before outreach, signup, or routing. Required: email.

util_ends_with(text, suffix)

Check whether text ends with a suffix Required: text, suffix.

util_ensure_prefix(text, prefix)

Ensure a string starts with a required prefix without duplication Required: text, prefix.

util_ensure_suffix(text, suffix)

Ensure a string ends with a required suffix without duplication Required: text, suffix.

util_ensure_trailing_slash(path)

Ensure a path ends with / Required: path.

util_env_key_validate(key)

Validate environment variable key shape A-Z0-9 underscore Required: key.

util_error_budget(target_percent, total_requests, failed_requests)

Calculate remaining failure budget and burn from traffic, failures, and an SLO target. Required: total_requests, failed_requests.

util_error_budget_plan(slo_percent, window_requests)

Convert an SLO percentage and request window into allowed failure budget. Required: slo_percent, window_requests.

util_etag_comparison_mode(left, weak, right)

Etag Comparison Mode evaluates strong or weak comparison between two entity tags from bounded caller-supplied values without an external provider. Call Etag Comparison Mode before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Etag Comparison Mode as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: left, right, weak.

util_etag_match(etag, other)

Compare ETag validators including weak tags and If-None-Match lists Required: etag, other.

util_etag_parse(etag)

Etag Parse separates weak and strong entity-tag semantics from an ETag value from bounded caller-supplied values without an external provider. Call Etag Parse before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Etag Parse as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: etag.

util_eth_to_wei(eth)

Scale ether amounts into wei integer strings for transfers and tx value Required: eth.

util_evm_access_list_shape_check(access_list)

EVM Access List Shape Check checks bounded EIP-2930 access-list address and storage-key shapes from bounded caller-supplied values without an external provider. Call EVM Access List Shape Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Access List Shape Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: access_list.

util_evm_address_list_dedupe(addresses)

EVM Address List Dedupe deduplicates EVM address strings case-insensitively while preserving order from bounded caller-supplied values without an external provider. Call EVM Address List Dedupe before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Address List Dedupe as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: addresses.

util_evm_address_zero_check(address)

EVM Address Zero Check detects the all-zero EVM address after shape normalization from bounded caller-supplied values without an external provider. Call EVM Address Zero Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Address Zero Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: address.

util_evm_allowance_budget_check(required_raw, allowance_raw)

EVM Allowance Budget Check checks caller-supplied allowance against a required token amount from bounded caller-supplied values without an external provider. Call EVM Allowance Budget Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Allowance Budget Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: allowance_raw, required_raw.

util_evm_amount_min_check(actual_raw, minimum_raw)

EVM Amount Min Check checks an actual raw amount against a minimum raw amount from bounded caller-supplied values without an external provider. Call EVM Amount Min Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Amount Min Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: actual_raw, minimum_raw.

util_evm_approve_amount_check(maximum_raw, proposed_raw, required_raw)

EVM Approve Amount Check compares a proposed approval with an explicit spending requirement and cap from bounded caller-supplied values without an external provider. Call EVM Approve Amount Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Approve Amount Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: proposed_raw, required_raw, maximum_raw.

util_evm_base_fee_growth_ceiling(blocks, growth_fraction, current_base_fee_wei)

EVM Base Fee Growth Ceiling compounds a caller growth fraction across future blocks from bounded caller-supplied values without an external provider. Call EVM Base Fee Growth Ceiling before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Base Fee Growth Ceiling as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: current_base_fee_wei, growth_fraction, blocks.

util_evm_batch_duplicate_nonce_check(nonces)

EVM Batch Duplicate Nonce Check finds duplicate nonces inside a caller-supplied transaction batch from bounded caller-supplied values without an external provider. Call EVM Batch Duplicate Nonce Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Batch Duplicate Nonce Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: nonces.

util_evm_batch_gas_sum(transactions)

EVM Batch Gas Sum sums gas limits and fee ceilings across a transaction batch from bounded caller-supplied values without an external provider. Call EVM Batch Gas Sum before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Batch Gas Sum as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: transactions.

util_evm_batch_value_sum(values_wei)

EVM Batch Value Sum sums bounded raw transaction values for a batch funding view from bounded caller-supplied values without an external provider. Call EVM Batch Value Sum before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Batch Value Sum as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: values_wei.

util_evm_blob_count_check(maximum, blob_count)

EVM Blob Count Check checks caller-supplied blob count against a transaction limit from bounded caller-supplied values without an external provider. Call EVM Blob Count Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Blob Count Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: blob_count, maximum.

util_evm_blob_fee_budget(blob_gas, max_fee_per_blob_gas_wei)

EVM Blob Fee Budget calculates a worst-case blob fee from blob gas and fee cap from bounded caller-supplied values without an external provider. Call EVM Blob Fee Budget before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Blob Fee Budget as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: blob_gas, max_fee_per_blob_gas_wei.

util_evm_block_age_check(now_epoch, block_timestamp, maximum_age_seconds)

EVM Block Age Check checks caller-supplied block timestamp age against a freshness budget from bounded caller-supplied values without an external provider. Call EVM Block Age Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Block Age Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: block_timestamp, now_epoch, maximum_age_seconds.

util_evm_calldata_selector_extract(calldata)

EVM Calldata Selector Extract extracts the first four calldata bytes as a selector from bounded caller-supplied values without an external provider. Call EVM Calldata Selector Extract before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Calldata Selector Extract as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: calldata.

util_evm_calldata_shape_check(calldata)

EVM Calldata Shape Check checks 0x-prefixed calldata for even-length hexadecimal bytes from bounded caller-supplied values without an external provider. Call EVM Calldata Shape Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Calldata Shape Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: calldata.

util_evm_calldata_size_bytes(calldata)

EVM Calldata Size Bytes calculates calldata byte length from normalized hex from bounded caller-supplied values without an external provider. Call EVM Calldata Size Bytes before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Calldata Size Bytes as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: calldata.

util_evm_calldata_word_count(calldata)

EVM Calldata Word Count counts complete and partial 32-byte calldata words after the selector from bounded caller-supplied values without an external provider. Call EVM Calldata Word Count before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Calldata Word Count as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: calldata.

util_evm_chain_id_hex_normalize(chain_id)

EVM Chain Id Hex Normalize converts decimal or hexadecimal chain identifiers into both forms from bounded caller-supplied values without an external provider. Call EVM Chain Id Hex Normalize before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Chain Id Hex Normalize as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: chain_id.

util_evm_chain_match_check(expected_chain_id, transaction_chain_id)

EVM Chain Match Check compares transaction and expected chain identifiers from bounded caller-supplied values without an external provider. Call EVM Chain Match Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Chain Match Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: transaction_chain_id, expected_chain_id.

util_evm_confirmation_count(current_block, inclusion_block)

EVM Confirmation Count calculates confirmations from inclusion and current block numbers from bounded caller-supplied values without an external provider. Call EVM Confirmation Count before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Confirmation Count as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: inclusion_block, current_block.

util_evm_contract_creation_check(to)

EVM Contract Creation Check classifies an empty or absent to-address as contract creation from bounded caller-supplied values without an external provider. Call EVM Contract Creation Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Contract Creation Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: to.

util_evm_deadline_check(now_epoch, deadline_epoch, minimum_lead_seconds)

EVM Deadline Check checks a transaction deadline against current time and minimum lead time from bounded caller-supplied values without an external provider. Call EVM Deadline Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Deadline Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: deadline_epoch, now_epoch, minimum_lead_seconds.

util_evm_eip1559_fee_check(gas_limit, max_fee_per_gas_wei, max_priority_fee_per_gas_wei)

EVM Eip1559 Fee Check checks dynamic-fee caps and calculates the worst-case transaction fee from bounded caller-supplied values without an external provider. Call EVM Eip1559 Fee Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Eip1559 Fee Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: max_fee_per_gas_wei, max_priority_fee_per_gas_wei, gas_limit.

util_evm_event_log_shape_check(log)

EVM Event Log Shape Check checks address, topics, and data shapes for a caller-supplied log from bounded caller-supplied values without an external provider. Call EVM Event Log Shape Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Event Log Shape Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: log.

util_evm_fee_cap_order_check(max_fee_per_gas_wei, max_priority_fee_per_gas_wei)

EVM Fee Cap Order Check checks that the priority fee cap does not exceed the total fee cap from bounded caller-supplied values without an external provider. Call EVM Fee Cap Order Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Fee Cap Order Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: max_fee_per_gas_wei, max_priority_fee_per_gas_wei.

util_evm_finality_depth_check(current_block, inclusion_block, required_confirmations)

EVM Finality Depth Check checks confirmation count against a caller-required depth from bounded caller-supplied values without an external provider. Call EVM Finality Depth Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Finality Depth Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: inclusion_block, current_block, required_confirmations.

util_evm_gas_cost_ceiling(gas_limit, max_fee_per_gas_wei)

EVM Gas Cost Ceiling calculates a worst-case native fee from gas limit and fee cap from bounded caller-supplied values without an external provider. Call EVM Gas Cost Ceiling before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Gas Cost Ceiling as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: gas_limit, max_fee_per_gas_wei.

util_evm_gas_limit_headroom(gas_limit, estimated_gas)

EVM Gas Limit Headroom calculates gas-limit headroom above an estimated gas amount from bounded caller-supplied values without an external provider. Call EVM Gas Limit Headroom before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Gas Limit Headroom as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: estimated_gas, gas_limit.

util_evm_hex_data_normalize(data)

EVM Hex Data Normalize normalizes even-length EVM hex data while preserving byte boundaries from bounded caller-supplied values without an external provider. Call EVM Hex Data Normalize before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Hex Data Normalize as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: data.

util_evm_hex_quantity_normalize(quantity)

EVM Hex Quantity Normalize normalizes an EVM hex quantity without redundant leading zeroes from bounded caller-supplied values without an external provider. Call EVM Hex Quantity Normalize before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Hex Quantity Normalize as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: quantity.

util_evm_infinite_approval_flag(approval_raw, threshold_raw)

EVM Infinite Approval Flag flags an approval at or near a caller-supplied maximum uint threshold from bounded caller-supplied values without an external provider. Call EVM Infinite Approval Flag before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Infinite Approval Flag as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: approval_raw, threshold_raw.

util_evm_legacy_fee_check(gas_limit, gas_price_wei, maximum_fee_wei)

EVM Legacy Fee Check checks gasPrice and gasLimit values for a legacy transaction budget from bounded caller-supplied values without an external provider. Call EVM Legacy Fee Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Legacy Fee Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: gas_price_wei, gas_limit, maximum_fee_wei.

util_evm_log_topic_count_check(topics, maximum)

EVM Log Topic Count Check checks log topic count against an expected bound from bounded caller-supplied values without an external provider. Call EVM Log Topic Count Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Log Topic Count Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: topics, maximum.

util_evm_multisend_recipient_check(allowed, recipients)

EVM Multisend Recipient Check checks duplicate, zero, and unexpected recipients in a multisend from bounded caller-supplied values without an external provider. Call EVM Multisend Recipient Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Multisend Recipient Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: recipients, allowed.

util_evm_native_balance_budget(value_wei, balance_wei, fee_ceiling_wei)

EVM Native Balance Budget checks caller-supplied native balance against value plus fee reserve from bounded caller-supplied values without an external provider. Call EVM Native Balance Budget before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Native Balance Budget as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: balance_wei, value_wei, fee_ceiling_wei.

util_evm_next_base_fee_bound(current_base_fee_wei, maximum_change_fraction)

EVM Next Base Fee Bound calculates upper and lower next-block base-fee bounds from bounded caller-supplied values without an external provider. Call EVM Next Base Fee Bound before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Next Base Fee Bound as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: current_base_fee_wei, maximum_change_fraction.

util_evm_nonce_gap_check(maximum_gap, pending_nonce, proposed_nonce)

EVM Nonce Gap Check calculates the gap between pending and proposed transaction nonces from bounded caller-supplied values without an external provider. Call EVM Nonce Gap Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Nonce Gap Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: pending_nonce, proposed_nonce, maximum_gap.

util_evm_nonce_replace_check(pending_nonces, proposed_nonce)

EVM Nonce Replace Check checks whether a proposed nonce targets a caller-supplied pending nonce from bounded caller-supplied values without an external provider. Call EVM Nonce Replace Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Nonce Replace Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: pending_nonces, proposed_nonce.

util_evm_priority_fee_ratio(max_fee_per_gas_wei, max_priority_fee_per_gas_wei)

EVM Priority Fee Ratio calculates the priority-fee share of the total fee cap from bounded caller-supplied values without an external provider. Call EVM Priority Fee Ratio before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Priority Fee Ratio as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: max_fee_per_gas_wei, max_priority_fee_per_gas_wei.

util_evm_receipt_gas_variance(gas_used, estimated_gas)

EVM Receipt Gas Variance compares gas used with a preflight estimate and reports variance from bounded caller-supplied values without an external provider. Call EVM Receipt Gas Variance before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Receipt Gas Variance as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: estimated_gas, gas_used.

util_evm_receipt_status_check(status)

EVM Receipt Status Check normalizes receipt status values into success or failure from bounded caller-supplied values without an external provider. Call EVM Receipt Status Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Receipt Status Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: status.

util_evm_replacement_fee_bump(new_fee_wei, old_fee_wei, minimum_bump_fraction)

EVM Replacement Fee Bump checks proposed fee caps against a required replacement bump from bounded caller-supplied values without an external provider. Call EVM Replacement Fee Bump before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Replacement Fee Bump as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: old_fee_wei, new_fee_wei, minimum_bump_fraction.

util_evm_signature_rs_shape_check(r, s)

EVM Signature Rs Shape Check checks r and s values as 32-byte hexadecimal scalars from bounded caller-supplied values without an external provider. Call EVM Signature Rs Shape Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Signature Rs Shape Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: r, s.

util_evm_signature_v_classify(v, chain_id)

EVM Signature V Classify classifies common signature recovery and EIP-155 v values from bounded caller-supplied values without an external provider. Call EVM Signature V Classify before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Signature V Classify as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: v, chain_id.

util_evm_slippage_bounds_check(minimum_output_raw, expected_output_raw, maximum_slippage_bps)

EVM Slippage Bounds Check checks expected and minimum output against a maximum slippage policy from bounded caller-supplied values without an external provider. Call EVM Slippage Bounds Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Slippage Bounds Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: expected_output_raw, minimum_output_raw, maximum_slippage_bps.

util_evm_token_balance_budget(amount_raw, balance_raw)

EVM Token Balance Budget checks caller-supplied raw token balance against transfer amount from bounded caller-supplied values without an external provider. Call EVM Token Balance Budget before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Token Balance Budget as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: balance_raw, amount_raw.

util_evm_transfer_recipient_check(encoded_recipient, intended_recipient)

EVM Transfer Recipient Check compares an encoded recipient address with the intended recipient from bounded caller-supplied values without an external provider. Call EVM Transfer Recipient Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Transfer Recipient Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: encoded_recipient, intended_recipient.

util_evm_tx_envelope_check(fields, transaction_type)

EVM Tx Envelope Check checks required envelope fields for a declared transaction type from bounded caller-supplied values without an external provider. Call EVM Tx Envelope Check before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Tx Envelope Check as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: transaction_type, fields.

util_evm_tx_type_classify(transaction_type)

EVM Tx Type Classify classifies legacy, access-list, dynamic-fee, and blob transaction type codes from bounded caller-supplied values without an external provider. Call EVM Tx Type Classify before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Tx Type Classify as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: transaction_type.

util_evm_value_plus_fee_budget(value_wei, fee_ceiling_wei)

EVM Value Plus Fee Budget sums transaction value and fee ceiling for a funding requirement from bounded caller-supplied values without an external provider. Call EVM Value Plus Fee Budget before signing, submitting, replacing, or accepting an EVM/Base transaction assembled from caller-owned data. Returns normalized transaction evidence, calculated budget or shape findings, and a fail-closed preflight status for EVM Value Plus Fee Budget as versioned deterministic JSON. Limitation: Uses caller-supplied values only; it performs no RPC call, signing, broadcasting, balance lookup, custody, trading, or live-chain claim. Required: value_wei, fee_ceiling_wei.

util_exec_quality_vs_twap(side, twap, avg_fill)

Compare average fill to TWAP benchmark (bps) Required: avg_fill, twap.

util_exec_quality_vs_vwap(side, vwap, avg_fill)

Compare average fill to VWAP benchmark (bps) Required: avg_fill, vwap.

util_expectancy(pnls)

Expectancy per trade from win rate, avg win, and avg loss Required: pnls.

util_expected_log_growth(fraction, true_prob, market_price)

Approx expected log bankroll growth for a binary bet fraction Required: true_prob, market_price, fraction.

util_experiment_lift(control_visitors, treatment_visitors, control_conversions, treatment_conversions)

Calculate descriptive conversion lift for control and treatment. Required: control_conversions, control_visitors, treatment_conversions, treatment_visitors.

util_expires_header_validate(date_epoch, expires_epoch)

Expires Header Validate compares Date and Expires timestamps for an explicit cache lifetime from bounded caller-supplied values without an external provider. Call Expires Header Validate before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Expires Header Validate as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: date_epoch, expires_epoch.

util_ext_from_mime(mime)

Guess extension from MIME type Required: mime.

util_extract_emails

Extract email-like addresses from text (regex only). Required: text.

util_extract_emails_lite(text)

Extract email-shaped tokens from free text (shape only) Required: text.

util_extract_ints(text)

Extract signed integers from free text Required: text.

util_extract_numbers(text)

Delx first-party micro-utility util_extract_numbers at $0.001 USDC. Required: text.

util_extract_urls

Extract HTTP/HTTPS URLs from text. Required: text.

util_factorial(n)

Delx first-party micro-utility util_factorial at $0.001 USDC. Required: n.

util_fee_quote(fee_rate, notional)

Fee amount = notional × fee_rate for maker/taker style quotes Required: notional, fee_rate.

util_feed_discover(url, timeout)

Find RSS, Atom, and JSON feeds so agents can subscribe instead of scrape. Required: url.

util_fibonacci(n)

Delx first-party micro-utility util_fibonacci at $0.001 USDC. Required: n.

util_field_coverage(records)

Measure per-field presence across bounded JSON records. Required: records.

util_fingerprint_lite(text)

16-hex fingerprint of text via SHA-256 prefix Required: text.

util_flatten1(items)

Flatten one level of nested lists Required: items.

util_fleet_status_onepager_stub(agents)

Build a compact fleet status markdown from agent rows Required: agents.

util_floor_n(value)

Floor a number to integer Required: value.

util_forms_extract(url, timeout)

Extract forms, methods, actions, and fields for browser automation and workflow planning. Required: url.

util_forwarded_header_parse(header)

Forwarded Header Parse parses bounded standardized Forwarded header elements from bounded caller-supplied values without an external provider. Call Forwarded Header Parse before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Forwarded Header Parse as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header.

util_fraction_to_bps(fraction)

Convert fractional rates into basis points for trading UIs Required: fraction.

util_frame_ancestors_check(header, required_source)

Frame Ancestors Check checks whether a CSP frame-ancestors directive contains a required source from bounded caller-supplied values without an external provider. Call Frame Ancestors Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Frame Ancestors Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header, required_source.

util_funding_annualize(hours, funding_rate)

Annualize a period funding rate (e.g. 8h → APR) Required: funding_rate, hours.

util_funding_payment(hours, notional, funding_rate)

Funding payment ≈ notional × funding_rate for a period Required: notional, funding_rate, hours.

util_funnel_conversion(stages)

Calculate step and overall conversion from ordered funnel stages. Required: stages.

util_fx_convert_amount(base, rate, side, quote, amount)

Convert an amount with a caller-supplied FX rate (no live feed) Required: amount, rate.

util_fx_revalue_amount(amount, rate_to, rate_from)

Revalue amount by two rates (historical vs current) Required: amount, rate_from, rate_to.

util_gas_cost_eth(gas_used, gas_price_wei)

Compute total gas cost in ether from used gas and price wei Required: gas_used, gas_price_wei.

util_gas_cost_usd(eth_usd, gas_used, gas_price_wei)

Estimate gas cost in USD given used gas, price, and ETH/USD Required: gas_used, gas_price_wei, eth_usd.

util_gas_cost_wei(gas_used, gas_price_wei)

Compute gas_used × gas_price as total gas cost in wei Required: gas_used, gas_price_wei.

util_gcd(a, b)

Delx first-party micro-utility util_gcd at $0.001 USDC. Required: a, b.

util_gdelt_query_normalize(q)

Normalize a news query string for GDELT-style search Required: q.

util_graphql_operation_index(document)

Index named GraphQL queries, mutations, and subscriptions. Required: document.

util_gross_margin(cost, price)

Calculate gross profit, gross margin, and profitability. Required: price, cost.

util_gwei_to_wei(gwei)

Convert gas prices from gwei into wei integer strings for EVM transactions Required: gwei.

util_half_kelly(b, win_prob)

Half-Kelly fraction for conservative bankroll growth Required: win_prob, b.

util_handoff_size_estimate(text)

Estimate bytes/tokens for a handoff capsule string Required: text.

util_hash(input, algorithm)

Hash a string with SHA-256, SHA-1, or MD5. Required: input.

util_header_map_normalize(headers)

Lowercase HTTP header keys and drop hop-by-hop headers Required: headers.

util_hex_canonicalize(hex, prefer_0x)

Canonicalize and validate even-length hex strings with optional 0x prefix. Required: hex.

util_hex_color_luminance(hex)

Relative luminance of a #RRGGBB color Required: hex.

util_hex_convert

Encode text to hex or decode hex to text. Required: text.

util_hex_prefix(hex, prefer_0x)

Normalize hex with optional 0x prefix and even-length check Required: hex.

util_hex_to_int(hex)

Parse 0x-hex strings into decimal integer strings for agents Required: hex.

util_hex_to_rgb(hex)

Convert #rrggbb to RGB channels Required: hex.

util_hmac_sha256_hex(text, secret)

HMAC-SHA256 hex of text with a secret Required: text, secret.

util_holidays(year, calendar)

List observed US federal holidays for a year. Required: year.

util_host_port_split(host)

Split host and optional port from a host:port or URL authority string Required: host.

util_hsts_policy_check(header, minimum_seconds, require_subdomains)

HSTS Policy Check checks HSTS max-age and subdomain directives against caller policy from bounded caller-supplied values without an external provider. Call HSTS Policy Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for HSTS Policy Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header, minimum_seconds, require_subdomains.

util_html_canonical_link_check(document_url, canonical_url)

Html Canonical Link Check compares document and canonical URLs after deterministic normalization from bounded caller-supplied values without an external provider. Call Html Canonical Link Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Html Canonical Link Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: document_url, canonical_url.

util_html_escape(text)

HTML-escape text for safe embedding Required: text.

util_html_meta_refresh_parse(content, base_url)

Html Meta Refresh Parse parses bounded meta-refresh delay and target URL from bounded caller-supplied values without an external provider. Call Html Meta Refresh Parse before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Html Meta Refresh Parse as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: content, base_url.

util_html_strip(html, max_length)

Strip HTML tags to plain text (best-effort, no browser). Required: html.

util_html_unescape(text)

HTML-unescape entities in text Required: text.

util_http_codes(code)

Look up HTTP status codes. Returns name, description, and category. Without code param, returns common codes.

util_http_header_diff(after, before)

Compare HTTP header maps case-insensitively while redacting credential-bearing values. Required: before, after.

util_http_headers_inspect(url, timeout)

Inspect security, cache, redirect, and server headers to audit a URL quickly. Required: url.

util_http_request_canonicalize(request)

Canonicalize an HTTP request, remove credentials, and derive a stable fingerprint. Required: request.

util_http_retry_plan(method, idempotent, status_code)

Decide whether an HTTP failure is safe to retry and return an idempotency-aware policy. Required: status_code.

util_http_version_normalize(version)

HTTP Version Normalize normalizes common HTTP protocol version spellings from bounded caller-supplied values without an external provider. Call HTTP Version Normalize before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for HTTP Version Normalize as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: version.

util_iban_checksum(iban)

Validate IBAN checksum (mod 97) offline Required: iban.

util_idempotency_fingerprint(scope, payload)

Create a stable SHA-256 idempotency fingerprint from a scope and canonical payload without storing state. Required: payload.

util_idempotency_key(payload, namespace)

Derive a deterministic namespaced idempotency key from canonical JSON. Required: payload.

util_il_constant_product(price_ratio)

Classic CPMM impermanent loss from price ratio Required: price_ratio.

util_il_fee_breakeven(price_ratio)

Fees needed to offset IL at a price ratio Required: price_ratio.

util_image_prompt_lint(prompt, max_chars)

Lint a text-to-image prompt for length and basic policy tokens before paying Required: prompt.

util_image_result_contract_check(result, max_width, max_height, max_size_bytes)

Validate image delivery metadata without fetching, transforming, or generating media. Required: result.

util_impact_price(mid, side, slippage_bps)

Expected impact price from mid and signed slippage fraction Required: mid, slippage_bps, side.

util_implied_from_yes_no(no, yes)

Normalize yes/no prices into implied probabilities Required: yes, no.

util_in_range(max, min, value)

Check whether a value sits in [min,max] Required: value, min, max.

util_incident_severity(impact)

Classify operational incident severity from explicit impact signals. Required: impact.

util_incident_timeline(events)

Normalize and chronologically order incident events into a machine-readable timeline. Required: events.

util_indent_text(text, spaces)

Indent every line by N spaces Required: text.

util_index_of(text, needle)

Find first index of a substring Required: text, needle.

util_inflation_calculator(amount, series, to_year, from_year)

Revalue an amount between two years using U.S. BLS CPI (CPI-U default). Informational only. Required: amount, from_year, to_year.

util_initials(text)

Delx first-party micro-utility util_initials at $0.001 USDC. Required: text.

util_int_to_hex(value)

Encode non-negative integers into 0x-hex strings Required: value.

util_int_to_roman(n)

Delx first-party micro-utility util_int_to_roman at $0.001 USDC. Required: n.

util_inventory_skew(target, inventory)

Skew an inventory target away from neutral for MM heuristics Required: inventory, target.

util_iota(n)

Generate 0..n-1 integers Required: n.

util_ip_classify(ip)

Classify an IPv4/IPv6 address as private, loopback, global, etc. No geolocation. Required: ip.

util_ip_v4_private_check(ip)

Detect RFC1918 private IPv4 addresses Required: ip.

util_ipv4_octets(ip)

Parse IPv4 into octets Required: ip.

util_ipv4_private(ip)

Detect private/loopback IPv4 Required: ip.

util_is_blank(text)

Detect blank or whitespace-only strings before paying downstream Required: text.

util_is_email_shape(email)

Validate email shape (not mailbox existence) Required: email.

util_is_empty_obj(value)

Check empty object from value Required: value.

util_is_float_str(text)

Check whether a string is a float Required: text.

util_is_hex_shape(hex)

Validate even-length hex (optional 0x) Required: hex.

util_is_int_str(text)

Check whether a string is a signed integer Required: text.

util_is_json_arr(value)

Check whether value is a JSON array Required: value.

util_is_json_obj(value)

Check whether value is a JSON object Required: value.

util_is_leap_year(year)

Delx first-party micro-utility util_is_leap_year at $0.001 USDC. Required: year.

util_is_prime(n)

Delx first-party micro-utility util_is_prime at $0.001 USDC. Required: n.

util_is_subset(a, b)

Check whether list a is subset of list b Required: a, b.

util_is_url_shape(url)

Validate http(s) URL shape Required: url.

util_is_uuid_shape(uuid)

Validate UUID string shape Required: uuid.

util_is_weekend(date)

Delx first-party micro-utility util_is_weekend at $0.001 USDC. Required: date.

util_isbn_check(isbn)

Validate ISBN-10 or ISBN-13 checksums. Required: isbn.

util_iso_duration_parse(duration)

Parse ISO-8601 duration subset PnDTnHnMnS into total seconds Required: duration.

util_iso_to_unix(iso)

Convert ISO-8601 to unix seconds Required: iso.

util_iso_week(date)

Delx first-party micro-utility util_iso_week at $0.001 USDC. Required: date.

util_iso3166_country_lookup(code)

Lookup country name from a static ISO3166-1 alpha-2 table subset Required: code.

util_join_csv(fields)

Join fields into a CSV row with proper quoting Required: fields.

util_join_lines(lines, separator)

Join an array of lines with a separator Required: lines.

util_json_array_chunk(items, chunk_size)

JSON Array Chunk chunks a JSON array into bounded ordered groups from bounded caller-supplied values without an external provider. Call JSON Array Chunk before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Array Chunk as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: items, chunk_size.

util_json_array_dedupe(items)

JSON Array Dedupe removes duplicate JSON array values while preserving first occurrence from bounded caller-supplied values without an external provider. Call JSON Array Dedupe before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Array Dedupe as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: items.

util_json_array_difference(left, right)

JSON Array Difference returns left-side values absent from the right JSON array from bounded caller-supplied values without an external provider. Call JSON Array Difference before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Array Difference as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: left, right.

util_json_array_filter_keys(key, items, allowed)

JSON Array Filter Keys keeps object rows whose selected key matches caller values from bounded caller-supplied values without an external provider. Call JSON Array Filter Keys before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Array Filter Keys as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: items, key, allowed.

util_json_array_flatten_one(items)

JSON Array Flatten One flattens exactly one nesting level from a JSON array from bounded caller-supplied values without an external provider. Call JSON Array Flatten One before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Array Flatten One as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: items.

util_json_array_group_by_key(key, items)

JSON Array Group By Key groups object rows by a selected top-level key from bounded caller-supplied values without an external provider. Call JSON Array Group By Key before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Array Group By Key as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: items, key.

util_json_array_index_by_key(key, items)

JSON Array Index By Key indexes object rows by a unique top-level key and reports duplicates from bounded caller-supplied values without an external provider. Call JSON Array Index By Key before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Array Index By Key as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: items, key.

util_json_array_intersection(left, right)

JSON Array Intersection returns values shared by two JSON arrays in left-side order from bounded caller-supplied values without an external provider. Call JSON Array Intersection before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Array Intersection as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: left, right.

util_json_array_length_check(fields, object, maximum, minimum)

JSON Array Length Check checks selected array fields against caller item-count bounds from bounded caller-supplied values without an external provider. Call JSON Array Length Check before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Array Length Check as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object, fields, minimum, maximum.

util_json_array_project_keys(keys, items)

JSON Array Project Keys projects selected keys from every object row from bounded caller-supplied values without an external provider. Call JSON Array Project Keys before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Array Project Keys as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: items, keys.

util_json_array_sort_by_key(key, items, descending)

JSON Array Sort By Key sorts object rows by a selected top-level key from bounded caller-supplied values without an external provider. Call JSON Array Sort By Key before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Array Sort By Key as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: items, key, descending.

util_json_boolean_coerce(values)

JSON Boolean Coerce coerces explicit boolean tokens while reporting rejected values from bounded caller-supplied values without an external provider. Call JSON Boolean Coerce before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Boolean Coerce as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: values.

util_json_contract_check(contract, instance)

Check a JSON payload against a bounded structural contract and return path-level errors. Required: instance, contract.

util_json_cross_field_equal(object, left_field, right_field)

JSON Cross Field Equal checks two top-level fields for exact JSON equality from bounded caller-supplied values without an external provider. Call JSON Cross Field Equal before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Cross Field Equal as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object, left_field, right_field.

util_json_cross_field_order(object, relation, left_field, right_field)

JSON Cross Field Order checks two numeric fields for a caller-selected ordering relation from bounded caller-supplied values without an external provider. Call JSON Cross Field Order before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Cross Field Order as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object, left_field, right_field, relation.

util_json_depth_check(value, maximum_depth)

JSON Depth Check measures JSON nesting depth against a caller budget from bounded caller-supplied values without an external provider. Call JSON Depth Check before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Depth Check as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: value, maximum_depth.

util_json_email_fields_check(fields, object)

JSON Email Fields Check checks selected fields for bounded email address shape from bounded caller-supplied values without an external provider. Call JSON Email Fields Check before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Email Fields Check as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object, fields.

util_json_empty_prune(object)

JSON Empty Prune removes selected empty string, array, and object top-level values from bounded caller-supplied values without an external provider. Call JSON Empty Prune before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Empty Prune as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object.

util_json_enum_fields_check(enums, object)

JSON Enum Fields Check checks selected fields against caller-provided enum values from bounded caller-supplied values without an external provider. Call JSON Enum Fields Check before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Enum Fields Check as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object, enums.

util_json_escape(text)

JSON-escape a string (without surrounding quotes) Required: text.

util_json_hash_fields_check(fields, object, hex_length)

JSON Hash Fields Check checks selected fields for caller-specified hexadecimal digest length from bounded caller-supplied values without an external provider. Call JSON Hash Fields Check before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Hash Fields Check as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object, fields, hex_length.

util_json_key_allowlist(object, allowed)

JSON Key Allowlist finds top-level object keys outside an explicit allowlist from bounded caller-supplied values without an external provider. Call JSON Key Allowlist before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Key Allowlist as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object, allowed.

util_json_key_denylist(denied, object)

JSON Key Denylist finds top-level object keys present in an explicit denylist from bounded caller-supplied values without an external provider. Call JSON Key Denylist before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Key Denylist as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object, denied.

util_json_keys(json_text)

Delx first-party micro-utility util_json_keys at $0.001 USDC. Required: json_text.

util_json_lines_parse(text)

JSON Lines Parse parses bounded newline-delimited JSON into ordered values from bounded caller-supplied values without an external provider. Call JSON Lines Parse before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Lines Parse as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: text.

util_json_lines_stringify(items)

JSON Lines Stringify serializes bounded JSON values as compact newline-delimited JSON from bounded caller-supplied values without an external provider. Call JSON Lines Stringify before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Lines Stringify as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: items.

util_json_merge_patch(patch, target)

Apply bounded top-level JSON Merge Patch semantics for integration planning. Required: target, patch.

util_json_minify(json_text)

Delx first-party micro-utility util_json_minify at $0.001 USDC. Required: json_text.

util_json_null_prune(object)

JSON Null Prune removes null-valued top-level keys from a JSON object from bounded caller-supplied values without an external provider. Call JSON Null Prune before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Null Prune as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object.

util_json_number_coerce(values)

JSON Number Coerce coerces explicit numeric strings while reporting rejected values from bounded caller-supplied values without an external provider. Call JSON Number Coerce before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Number Coerce as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: values.

util_json_numeric_range_check(fields, object, maximum, minimum)

JSON Numeric Range Check checks selected numeric fields against inclusive bounds from bounded caller-supplied values without an external provider. Call JSON Numeric Range Check before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Numeric Range Check as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object, fields, minimum, maximum.

util_json_object_deep_merge(left, right)

JSON Object Deep Merge deep-merges two bounded JSON objects with right-side precedence from bounded caller-supplied values without an external provider. Call JSON Object Deep Merge before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Object Deep Merge as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: left, right.

util_json_object_diff(after, before)

JSON Object Diff reports added, removed, and changed top-level object keys from bounded caller-supplied values without an external provider. Call JSON Object Diff before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Object Diff as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: before, after.

util_json_object_lowercase_keys(object)

JSON Object Lowercase Keys lowercases top-level keys and reports case collisions from bounded caller-supplied values without an external provider. Call JSON Object Lowercase Keys before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Object Lowercase Keys as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object.

util_json_object_omit(keys, object)

JSON Object Omit removes selected top-level keys from one JSON object from bounded caller-supplied values without an external provider. Call JSON Object Omit before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Object Omit as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object, keys.

util_json_object_pick(keys, object)

JSON Object Pick keeps selected top-level keys from one JSON object from bounded caller-supplied values without an external provider. Call JSON Object Pick before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Object Pick as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object, keys.

util_json_object_rename_keys(object, mapping)

JSON Object Rename Keys renames top-level keys using a caller mapping and reports collisions from bounded caller-supplied values without an external provider. Call JSON Object Rename Keys before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Object Rename Keys as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object, mapping.

util_json_object_sort_keys(object)

JSON Object Sort Keys returns a top-level JSON object with deterministic key order from bounded caller-supplied values without an external provider. Call JSON Object Sort Keys before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Object Sort Keys as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object.

util_json_patch_preview(document, operations)

JSON Patch Preview previews bounded add, remove, and replace operations without mutating a source from bounded caller-supplied values without an external provider. Call JSON Patch Preview before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Patch Preview as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: document, operations.

util_json_path_pick(path, object)

Pick a dotted path from a JSON object Required: object, path.

util_json_pattern_fields_check(fields, object, pattern)

JSON Pattern Fields Check checks selected string fields against a bounded regular expression from bounded caller-supplied values without an external provider. Call JSON Pattern Fields Check before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Pattern Fields Check as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object, fields, pattern.

util_json_pointer_escape(token)

JSON Pointer Escape escapes one JSON Pointer reference token from bounded caller-supplied values without an external provider. Call JSON Pointer Escape before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Pointer Escape as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: token.

util_json_pointer_get(pointer, document)

JSON Pointer Get resolves a bounded RFC 6901-style JSON pointer from bounded caller-supplied values without an external provider. Call JSON Pointer Get before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Pointer Get as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: document, pointer.

util_json_pointer_unescape(token)

JSON Pointer Unescape unescapes one JSON Pointer reference token from bounded caller-supplied values without an external provider. Call JSON Pointer Unescape before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Pointer Unescape as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: token.

util_json_pretty(json_text)

Delx first-party micro-utility util_json_pretty at $0.001 USDC. Required: json_text.

util_json_required_fields_check(object, required_fields)

JSON Required Fields Check finds absent or null required top-level fields from bounded caller-supplied values without an external provider. Call JSON Required Fields Check before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Required Fields Check as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object, required_fields.

util_json_schema_quickcheck(schema, instance)

Lightweight type/required/property check for agent-held JSON (not full draft validation) Required: instance, schema.

util_json_schema_validate_lite(schema, instance)

Lite validate object required keys and types vs a tiny schema Required: instance, schema.

util_json_shape_profile(data)

Profile the top-level shape, fields, size, and cardinality of a JSON value. Required: data.

util_json_size_budget(value, maximum_bytes)

JSON Size Budget measures compact UTF-8 JSON size against a byte budget from bounded caller-supplied values without an external provider. Call JSON Size Budget before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Size Budget as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: value, maximum_bytes.

util_json_string_length_check(fields, object, maximum, minimum)

JSON String Length Check checks selected string fields against caller length bounds from bounded caller-supplied values without an external provider. Call JSON String Length Check before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON String Length Check as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object, fields, minimum, maximum.

util_json_timestamp_fields_check(fields, object, maximum_epoch, minimum_epoch)

JSON Timestamp Fields Check checks selected fields as integer epoch seconds within bounds from bounded caller-supplied values without an external provider. Call JSON Timestamp Fields Check before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Timestamp Fields Check as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object, fields, minimum_epoch, maximum_epoch.

util_json_to_csv(delimiter, json_text)

Convert structured JSON rows into CSV for exports, spreadsheets, and handoff. Required: json_text.

util_json_type(json_text)

Delx first-party micro-utility util_json_type at $0.001 USDC. Required: json_text.

util_json_type_map_check(types, object)

JSON Type Map Check checks top-level fields against caller-declared JSON type names from bounded caller-supplied values without an external provider. Call JSON Type Map Check before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Type Map Check as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object, types.

util_json_type_name(value)

JSON type name for a value Required: value.

util_json_unique_fields_check(items, fields)

JSON Unique Fields Check finds duplicate composite keys across object rows from bounded caller-supplied values without an external provider. Call JSON Unique Fields Check before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON Unique Fields Check as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: items, fields.

util_json_url_fields_check(fields, object)

JSON URL Fields Check checks selected fields for absolute HTTP or HTTPS URL shape from bounded caller-supplied values without an external provider. Call JSON URL Fields Check before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON URL Fields Check as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object, fields.

util_json_uuid_fields_check(fields, object)

JSON UUID Fields Check checks selected fields for canonical UUID text shape from bounded caller-supplied values without an external provider. Call JSON UUID Fields Check before an agent hands structured data or a batch plan to a downstream tool with a stricter contract. Returns the bounded transformation or validation finding with counts, normalized values, and a versioned result contract for JSON UUID Fields Check as versioned deterministic JSON. Limitation: Processes bounded caller-supplied data in memory only; it performs no file fetch, database write, code execution, or retention. Required: object, fields.

util_json_validate(input)

Validate and pretty-print JSON. Returns validity, errors, and formatted output. Required: input.

util_jsonld_key_list(data)

List top-level keys from JSON-LD or JSON objects the caller already holds. Required: data.

util_jwt_claims_audit(claims, now_epoch)

Audit JWT claim names, expiry, and identifiers at an explicit reference time without returning claim values. Required: claims, now_epoch.

util_jwt_claims_humanize(jwt)

Decode JWT payload claims without signature verify Required: jwt.

util_jwt_exp_seconds(token)

Read JWT exp claim seconds remaining without verifying signature Required: token.

util_jwt_inspect(token)

Decode JWT claims quickly for auth debugging, routing, and token inspection. Required: token.

util_kb_to_bytes(kib)

Convert kibibytes to bytes Required: kib.

util_kebab_case(text)

Convert free text or identifiers to kebab-case slugs Required: text.

util_kebab_to_camel(text)

Convert kebab-case identifiers to lowerCamelCase Required: text.

util_kelly_binary(true_prob, market_price)

Kelly fraction for a binary market given edge and price Required: true_prob, market_price.

util_kelly_fraction(b, win_prob)

Kelly fraction from win probability and net payoff odds b Required: win_prob, b.

util_keyword_density(text, keywords)

Measure keyword density inside caller-supplied text. Required: text, keywords.

util_ksuid

Delx first-party micro-utility util_ksuid at $0.001 USDC.

util_language_script_detect(text)

Detect dominant Unicode script families in a text sample. Required: text.

util_last_modified_compare(after_epoch, before_epoch)

Last Modified Compare orders two Last-Modified instants and reports their delta from bounded caller-supplied values without an external provider. Call Last Modified Compare before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Last Modified Compare as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: before_epoch, after_epoch.

util_latency_sla_check(sla_ms, observed_ms)

Compare observed latency against an SLA threshold in milliseconds. Required: observed_ms, sla_ms.

util_latency_summary(latencies_ms)

Calculate p50, p95, p99, mean, min, and max from bounded latency samples. Required: latencies_ms.

util_lcm(a, b)

Delx first-party micro-utility util_lcm at $0.001 USDC. Required: a, b.

util_lei_lookup(lei)

Look up a 20-character Legal Entity Identifier (LEI) via GLEIF: legal name, status, jurisdiction, registration. Required: lei.

util_lerp(a, b, t)

Linear interpolation between a and b by t Required: a, b, t.

util_levenshtein

Compute Levenshtein edit distance and similarity. Required: a, b.

util_leverage_notional(margin, leverage)

Notional exposure from margin and leverage multiplier Required: margin, leverage.

util_line_count(text, keep_empty)

Count lines in caller-held text with nonempty breakdown Required: text.

util_link_header_parse(header)

Link Header Parse parses bounded Link header targets and relation parameters from bounded caller-supplied values without an external provider. Call Link Header Parse before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Link Header Parse as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header.

util_links_extract(url, limit, timeout)

Map internal and external links on a page for crawling, routing, and site inspection. Required: url.

util_liq_cascade_steps(entry, steps, leverage, maintenance)

Advisory liq price ladder over leverage steps Required: entry, leverage, maintenance.

util_liq_price_long(mmr, entry, leverage)

Approximate long liquidation price from entry, leverage, and maintenance Required: entry, leverage, mmr.

util_liq_price_short(mmr, entry, leverage)

Approximate short liquidation price from entry, leverage, and maintenance Required: entry, leverage, mmr.

util_list_chunk(items)

Delx first-party micro-utility util_list_chunk at $0.001 USDC. Required: items.

util_list_len(items)

Return length of a JSON array or separator-split string list Required: items.

util_list_len_safe(items)

Safe length of a list/JSON array Required: items.

util_location_header_resolve(base_url, location)

Location Header Resolve resolves a relative Location value against a caller-supplied base URL from bounded caller-supplied values without an external provider. Call Location Header Resolve before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Location Header Resolve as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: base_url, location.

util_log_event_normalize(log_text)

Normalize JSON or common text logs into one stable event envelope. Required: log_text.

util_log_loss(prob, outcome)

Log loss (cross-entropy) for a probability vs binary outcome Required: prob, outcome.

util_login_surface_report(url, timeout)

Inspect auth surface signals like login forms, signup links, reset links, and security headers. Required: url.

util_logit(prob)

Logit transform log(p/(1−p)) for probability-space models Required: prob.

util_loyalty_reward_quote(currency, event_id, bonus_points, points_per_unit, purchase_amount, tier_multiplier, redemption_value_per_point)

Calculate loyalty points, reward value, effective rebate, and a ledger-ready credit instruction without storing customer state. Required: purchase_amount.

util_ltv_cac(cac, ltv)

Calculate LTV-to-CAC ratio with a bounded health heuristic. Required: ltv, cac.

util_luhn_check(number)

Validate a digit string with the Luhn checksum (does not prove a card is real). Required: number.

util_luhn_valid(number)

Luhn check for digit strings (cards/IDs) Required: number.

util_maker_taker_fee(notional, maker_bps, taker_bps)

Split fee drag into maker and taker components for a notional Required: notional, maker_bps, taker_bps.

util_map_range(value, in_max, in_min, out_max, out_min)

Map a value from one range into another Required: value, in_min, in_max, out_min, out_max.

util_margin_required(leverage, notional)

Margin required for a notional at a given leverage Required: notional, leverage.

util_markdown_heading_index(text)

Index markdown ATX headings from caller-supplied text only. Required: text.

util_markdown_to_text(markdown, max_length)

Strip markdown formatting to plain text (best-effort). Required: markdown.

util_market_roi(entry, resolved_yes)

ROI from entry price and resolution payout on a binary share Required: entry, resolved_yes.

util_mask_middle(text)

Mask the middle of a string keeping ends Required: text.

util_mask_string(text)

Delx first-party micro-utility util_mask_string at $0.001 USDC. Required: text.

util_max_loss(size, stop, entry)

Max loss dollars for size and stop distance Required: entry, stop, size.

util_max_n(values)

Maximum of a numeric list Required: values.

util_mb_to_bytes(mib)

Convert mebibytes to bytes Required: mib.

util_mcp_server_readiness_report(url, timeout)

Score an MCP server for initialize, tools/list, schema hygiene, manifest discovery, and agent usability. Required: url.

util_md5_hex(text)

MD5 hex digest of text Required: text.

util_mean(values)

Delx first-party micro-utility util_mean at $0.001 USDC. Required: values.

util_median(values)

Delx first-party micro-utility util_median at $0.001 USDC. Required: values.

util_merge_shallow(a, b)

Shallow-merge two objects (b wins) Required: a, b.

util_method_idempotency_classify(method)

Method Idempotency Classify classifies an HTTP method as safe, idempotent, or non-idempotent from bounded caller-supplied values without an external provider. Call Method Idempotency Classify before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Method Idempotency Classify as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: method.

util_mev_sandwich_upper(amount_in, reserve_in, reserve_out)

Upper-bound CPMM price impact loss for size vs reserves Required: reserve_in, reserve_out, amount_in.

util_mid_price(ask, bid)

Mid price = (bid + ask) / 2 for spread-aware marks Required: bid, ask.

util_mime_from_ext(ext)

Guess MIME type from file extension Required: ext.

util_mime_lookup(filename, extension)

Look up a common MIME type from a file extension or filename.

util_min_n(values)

Minimum of a numeric list Required: values.

util_month_name(iso)

Month name from ISO date Required: iso.

util_ms_to_hms(ms)

Convert milliseconds to HH:MM:SS Required: ms.

util_multi_outcome_kelly(probs, prices, bankroll)

Proportional Kelly-like stakes across exclusive outcomes Required: probs, prices, bankroll.

util_multi_outcome_normalize(prices)

Normalize multi-outcome prices into a probability simplex Required: prices.

util_mx_host_parse(mx)

Parse MX preference and host from an MX record string Required: mx.

util_nanoid

Delx first-party micro-utility util_nanoid at $0.001 USDC.

util_next_weekday(date)

Delx first-party micro-utility util_next_weekday at $0.001 USDC. Required: date.

util_nonce_hex(bytes)

Generate a random hex nonce

util_normalize_whitespace

Normalize whitespace in text. Required: text.

util_notional(size, price)

Notional value = mark price × size for sizing and risk Required: price, size.

util_null_distribution(records)

Measure null and missing-field distribution across JSON records. Required: records.

util_number_base

Convert a number between bases 2-36. Required: value.

util_numeric_outlier_scan(values, z_threshold)

Identify statistical outlier indexes in bounded numeric samples. Required: values.

util_oauth_scope_audit(required, requested)

Compare requested OAuth scopes with the minimum required scopes. Required: requested, required.

util_ob_depth_notional(levels)

Sum price*size depth notional for a side ladder Required: levels.

util_ob_imbalance(ask_size, bid_size)

Book imbalance (bid_size-ask_size)/(bid_size+ask_size) Required: bid_size, ask_size.

util_ob_mid_from_levels(ask, bid)

Compute mid price from bid/ask top-of-book levels Required: bid, ask.

util_obj_empty(object)

Check whether an object is empty Required: object.

util_obj_get(key, object, default)

Get object[key] with default Required: object, key.

util_obj_has(key, object)

Check whether an object has a key Required: object, key.

util_obj_pick_exists(keys, object)

Pick only keys that exist on an object Required: object, keys.

util_obj_size(object)

Count object keys Required: object.

util_odds_american_to_prob(odds)

Convert American odds into implied probability Required: odds.

util_odds_chain_convert(decimal)

Convert among decimal/american/fractional odds and implied prob Required: decimal.

util_odds_decimal_to_prob(odds)

Convert decimal odds into implied probability (1/odds) Required: odds.

util_odds_fractional_to_prob(den, num)

Convert fractional odds (num/den) into implied probability Required: num, den.

util_odds_ratio(prob)

Odds ratio p/(1−p) from a probability Required: prob.

util_odds_vig_strip(odds_a, odds_b)

Strip proportional vig from two-way decimal odds to fair probs Required: odds_a, odds_b.

util_ohlc_aggregate(prices)

Aggregate a bounded price list into open, high, low, and close. Required: prices.

util_omit_fields(keys, object)

Omit keys from an object Required: object, keys.

util_open_graph(url, timeout)

Extract Open Graph and Twitter card fields to preview how a URL will render in feeds and agents. Required: url.

util_open_redirect_audit(target, allowed_hosts)

Check a redirect target against an exact hostname allowlist. Required: target, allowed_hosts.

util_openapi_breaking_change(new_spec, old_spec)

Detect removed OpenAPI paths and methods before an agent deploys a client-breaking change. Required: old_spec, new_spec.

util_openapi_operation_index(spec)

Turn an OpenAPI document into a compact operation index for planning and tool selection. Required: spec.

util_openapi_path_diff(paths_a, paths_b)

Diff path keys between two OpenAPI path maps Required: paths_a, paths_b.

util_openapi_summary(url, timeout)

Summarize an OpenAPI document including title, version, paths, tags, and likely auth surface. Required: url.

util_ops_health_score(latency_score, saturation_percent, success_rate_percent)

Combine success rate, latency score, and saturation into an ops health band. Required: success_rate_percent, latency_score, saturation_percent.

util_option_payoff_points(legs, spots)

Payoff points for multi-leg options at sample spots Required: legs, spots.

util_ordinal(n)

Format an integer as an English ordinal (1st, 2nd, 3rd…) Required: n.

util_origin_header_normalize(origin)

Origin Header Normalize normalizes scheme, host, and effective port from an Origin value from bounded caller-supplied values without an external provider. Call Origin Header Normalize before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Origin Header Normalize as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: origin.

util_overround(probs)

Overround/vig from a set of mutually exclusive implied probs Required: probs.

util_p95_estimate(samples)

Estimate p95 latency from a bounded numeric sample. Required: samples.

util_pad_left(fill, text, width)

Left-pad a string to a target width Required: text, width.

util_pad_right(fill, text, width)

Right-pad a string to a target width Required: text, width.

util_pad_string(text)

Delx first-party micro-utility util_pad_string at $0.001 USDC. Required: text.

util_page_extract(url, timeout)

Turn any URL into clean page metadata and readable text for search, routing, and summarization. Required: url.

util_page_links_summary(html, limit)

Extract href links from caller-held HTML without network access Required: html.

util_page_meta_summary(html)

Extract title and common meta tags from caller-held HTML Required: html.

util_pagination_plan(response)

Infer a bounded cursor, link, or page pagination plan from a response envelope. Required: response.

util_pairs_hedge_ratio(x, y)

OLS hedge ratio beta of y on x (no intercept) Required: x, y.

util_pairs_zscore(spread)

Z-score of spread series for pairs trading Required: spread.

util_parlay_odds_decimal(odds)

Combine independent decimal odds into parlay decimal odds Required: odds.

util_parlay_prob_indep(probs)

Independent parlay probability as product of leg probabilities Required: probs.

util_parse_bool_list(items)

Coerce a list of tokens into booleans Required: items.

util_parse_bytes(text)

Delx first-party micro-utility util_parse_bytes at $0.001 USDC. Required: text.

util_parse_float_safe(text, default)

Parse float with default on failure Required: text.

util_parse_int_safe(text, default)

Parse int with default on failure Required: text.

util_password_policy_audit(policy)

Audit password policy for length, breached-secret checks, and MFA. Required: policy.

util_path_ext(path)

Extract file extension from a path or URL path segment Required: path.

util_path_join(parts)

Join URL or filesystem-like path segments without traversal escape Required: parts.

util_path_traversal_scan(path)

Evaluate an untrusted path for unsafe parent-segment signals without returning secret data. Required: path.

util_payload_size_check(payload, hard_limit_bytes, soft_limit_bytes)

Measure payload byte size against soft and hard ceilings before paying Required: payload.

util_payout_if_yes(shares)

Payout if YES resolves given shares held (\$1/share style) Required: shares.

util_pct_of(part, whole)

Compute part/whole as percent Required: part, whole.

util_pct_to_bps(pct)

Convert percent values into basis points (1% = 100 bps) Required: pct.

util_pdf_header_sniff(base64_prefix)

Sniff PDF magic and approximate size from base64 prefix Required: base64_prefix.

util_peg_deviation_bps(peg, price)

Peg deviation in bps from target (usually 1.0) Required: price.

util_peg_tw_deviation(peg, prices, weights)

Time-weighted absolute peg deviation over samples Required: prices, weights.

util_percent_change(to, from)

Compute absolute and percent change between two numbers. Required: from, to.

util_percent_clamp(max, min, value)

Clamp a percentage-like number into a min/max range Required: value.

util_percentile(values)

Delx first-party micro-utility util_percentile at $0.001 USDC. Required: values.

util_permissions_policy_parse(header)

Permissions Policy Parse parses Permissions-Policy feature allowlists into normalized entries from bounded caller-supplied values without an external provider. Call Permissions Policy Parse before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Permissions Policy Parse as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header.

util_perp_premium(mark, index)

Premium (mark-index)/index in bps Required: mark, index.

util_phone_digits(phone)

Delx first-party micro-utility util_phone_digits at $0.001 USDC. Required: phone.

util_phone_e164_normalize(phone, default_region)

Normalize phone digits toward E.164-like form Required: phone.

util_pick_fields(keys, object)

Pick a shallow whitelist of keys from a JSON object Required: object, keys.

util_pii_redact(text)

Redact common PII candidates before logs, prompts, or downstream tools receive the text. Required: text.

util_pii_risk_scan(text)

Count likely personal-data classes without echoing detected values. Required: text.

util_pluck(key, items)

Pluck a key from a list of objects Required: items, key.

util_pluralize_lite(word, count)

Lightweight English pluralization helper Required: word.

util_pnl_abs(exit, size, entry)

Absolute profit/loss from entry, exit, and position size Required: entry, exit, size.

util_pnl_pct(exit, entry)

Percent profit/loss from entry and exit prices Required: entry, exit.

util_pnl_with_fees(exit, size, entry, fee_rate)

Net PnL after entry and exit fee rates on a sized position Required: entry, exit, size, fee_rate.

util_position_size_risk(stop, entry, equity, risk_pct)

Position size from account risk budget and stop distance Required: equity, risk_pct, entry, stop.

util_position_value(size, price)

Mark a position to a given price for equity snapshots Required: size, price.

util_price_ceiling_check(buffer_pct, price_usdc, ceiling_usdc)

Check whether a published USDC price stays under a buyer ceiling Required: price_usdc, ceiling_usdc.

util_price_floor(cost, target_margin_percent)

Calculate the minimum price required for a target gross margin. Required: cost, target_margin_percent.

util_price_tick_normalize(ticks, decimals)

Normalize caller-supplied positive price ticks to fixed decimals. Required: ticks.

util_pricing_page_extract(url, timeout)

Extract pricing-page signals like plan names, free trial hints, CTA patterns, and sales routes. Required: url.

util_prime_factors(n)

Delx first-party micro-utility util_prime_factors at $0.001 USDC. Required: n.

util_prob_clamp(prob)

Clamp a probability into (eps, 1−eps) for stable odds/math Required: prob.

util_prob_to_odds_american(prob)

Convert a probability into American odds Required: prob.

util_prob_to_odds_decimal(prob)

Convert a probability into decimal odds Required: prob.

util_profit_factor(pnls)

Profit factor = gross wins / gross losses Required: pnls.

util_profit_if_yes(cost, shares)

Profit if YES resolves after cost basis on shares Required: shares, cost.

util_prompt_injection_scan(text)

Screen untrusted text for prompt-injection and tool-coercion signals before an agent consumes it. Required: text.

util_public_registry_id_shape(value)

Validate shape of common registry IDs (LEI/WMI/EIN-like) Required: value.

util_qr_payload_classify(payload)

Classify a decoded QR/barcode string (url/text/wifi heuristic) Required: payload.

util_quarter_of_year(date)

Delx first-party micro-utility util_quarter_of_year at $0.001 USDC. Required: date.

util_query_param_pick(url, keys)

Pick named query parameters from a URL or raw query string Required: url.

util_query_string(query, action, params)

Parse or build URL query strings.

util_query_string_parse(query)

Parse query pairs and redact credential-like keys without fetching. Required: query.

util_queue_wait_estimate(workers, service_ms, queue_depth)

Estimate queue wait from depth, mean service time, and workers. Required: queue_depth, service_ms.

util_rad_to_deg(radians)

Convert radians to degrees Required: radians.

util_random_choice

Pick one or more items at random without replacement. Required: items.

util_random_int

Generate cryptographically secure random integers.

util_random_uuid_v4

Generate a random RFC 4122 UUID v4 for agent correlation IDs

util_range_header_parse(header, resource_size)

Range Header Parse parses one bounded HTTP byte range into start and end offsets from bounded caller-supplied values without an external provider. Call Range Header Parse before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Range Header Parse as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header, resource_size.

util_range_list(step, stop, start)

Generate a Python-style range list Required: start, stop.

util_rate_limit_parse(headers)

Normalize common HTTP rate-limit and retry headers. Required: headers.

util_rate_limit_window_check(now_epoch, remaining, requested, reset_epoch)

Rate Limit Window Check checks remaining quota against reset timing and requested calls from bounded caller-supplied values without an external provider. Call Rate Limit Window Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Rate Limit Window Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: remaining, requested, reset_epoch, now_epoch.

util_ratio(a, b)

Compute a/b ratio safely Required: a, b.

util_rdap_lookup(domain, timeout)

Fetch registrar, status, and registration dates for trust, compliance, and domain ops. Required: domain.

util_reading_time

Estimate reading time from word count. Required: text.

util_realized_pnl_fifo(exit, entry, close_size)

Lite FIFO-style realized PnL from sequential buy/sell fills Required: entry, exit, close_size.

util_record_fingerprint(record)

Create a stable SHA-256 fingerprint from canonical JSON. Required: record.

util_redact_digits(text)

Replace digits with # for safe previews Required: text.

util_redact_emails(text)

Delx first-party micro-utility util_redact_emails at $0.001 USDC. Required: text.

util_redact_emails_lite(text)

Replace email-shaped tokens with a redaction marker Required: text.

util_redirect_chain_check(urls, max_hops)

Redirect Chain Check checks a caller-supplied redirect chain for loops, downgrade, and hop budget from bounded caller-supplied values without an external provider. Call Redirect Chain Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Redirect Chain Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: urls, max_hops.

util_referrer_policy_check(value, allowed)

Referrer Policy Check classifies a Referrer-Policy value against a caller allowlist from bounded caller-supplied values without an external provider. Call Referrer Policy Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Referrer Policy Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: value, allowed.

util_refund_policy_preflight(policy, purchase)

Evaluate simple refund eligibility flags against policy JSON Required: policy, purchase.

util_regex_test(text, flags, pattern)

Test a regex pattern against text. Returns matches, groups, and count. Required: pattern, text.

util_reliability_curve(bins, outcomes, forecasts)

Reliability diagram points (mean forecast vs hit rate per bin) Required: forecasts, outcomes.

util_remove_punctuation(text)

Delx first-party micro-utility util_remove_punctuation at $0.001 USDC. Required: text.

util_rename_key(object, to_key, from_key)

Rename a single object key Required: object, from_key, to_key.

util_repeat_str(text, count)

Repeat a string a bounded number of times Required: text, count.

util_repeat_string(text)

Delx first-party micro-utility util_repeat_string at $0.001 USDC. Required: text.

util_resolution_payout_matrix(holdings, winning_index)

Payout matrix for multi-outcome shares at resolution Required: holdings, winning_index.

util_response_contract_diff(actual, expected)

Compare expected and actual top-level response fields and value types. Required: expected, actual.

util_retention_policy_check(max_days, retention_days)

Compare declared retention with a maximum policy window. Required: retention_days, max_days.

util_retry_after_parse(value, now_epoch)

Retry After Parse normalizes Retry-After seconds or an HTTP-date epoch into a delay from bounded caller-supplied values without an external provider. Call Retry After Parse before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Retry After Parse as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: value, now_epoch.

util_retry_budget_plan(factor, max_attempts, max_delay_ms, base_delay_ms)

Build a bounded exponential retry delay schedule with a hard delay cap. Required: max_attempts, base_delay_ms.

util_retry_safe_classify(method, status_code, body_mutates, idempotency_key_present)

Classify whether a failed HTTP-style call is safe to auto-retry Required: method.

util_return_pct(end, start)

Compute percent return between start and end values. Required: start, end.

util_revenue_concentration(accounts)

Measure top-account and top-three revenue concentration without returning account values. Required: accounts.

util_reverse_string(text)

Delx first-party micro-utility util_reverse_string at $0.001 USDC. Required: text.

util_rfq_crypto_template(need, budget_usdc)

Fill a Demand Board RFQ template for crypto math needs Required: need.

util_rgb_to_hex(b, g, r)

Convert RGB channels to #rrggbb Required: r, g, b.

util_risk_parity_weights(returns)

Inverse-vol risk parity weights from return series Required: returns.

util_risk_reward(stop, take, entry)

Risk–reward ratio = reward / risk for stop/target geometry Required: entry, stop, take.

util_robots_allow_path(path, robots_txt, user_agent)

Check path allow/disallow against caller-held robots.txt rules Required: robots_txt, path.

util_robots_directive_merge(generic, agent_specific)

Robots Directive Merge merges generic and named-agent robots directives with specific precedence from bounded caller-supplied values without an external provider. Call Robots Directive Merge before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Robots Directive Merge as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: generic, agent_specific.

util_robots_inspect(url, timeout)

Read robots.txt rules and sitemap declarations before crawling or indexing a domain. Required: url.

util_robots_meta_parse(content)

Robots Meta Parse parses robots meta directives into normalized index and follow decisions from bounded caller-supplied values without an external provider. Call Robots Meta Parse before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Robots Meta Parse as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: content.

util_robots_path_budget(robots)

Count allow/disallow rules from robots.txt text Required: robots.

util_roi_on_margin(pnl, margin)

Return on margin for a realized or mark PnL Required: pnl, margin.

util_roman_to_int(roman)

Delx first-party micro-utility util_roman_to_int at $0.001 USDC. Required: roman.

util_rot13

Apply ROT13 to Latin letters in text. Required: text.

util_round_n(value, digits)

Round a number to N decimal places Required: value.

util_rrule_daily_next(n, from_ts, interval_days)

Next N daily fires from a DAILY RRULE-like interval Required: interval_days, from_ts, n.

util_runway(cash, monthly_burn)

Calculate static cash runway from cash and monthly burn. Required: cash, monthly_burn.

util_safe_div(a, b)

Divide a/b with ok flag when b!=0 Required: a, b.

util_saturation_ratio(capacity, in_flight)

Compute capacity saturation from in-flight work and capacity. Required: in_flight, capacity.

util_schema_inference(records)

Infer observed field types and nullability from bounded JSON records. Required: records.

util_score_clip(max, min, score)

Clip a score into an inclusive numeric range. Required: score.

util_seconds_to_hms(seconds)

Convert seconds to HH:MM:SS Required: seconds.

util_secret_exposure_scan(text)

Detect likely credentials and private keys without returning their values. Required: text.

util_security_headers_score(headers)

Score the presence of five high-value browser security headers. Required: headers.

util_security_surface_compose(dns_ok, tls_ok, headers_ok)

Compose DNS/TLS/header signal booleans into a surface score Required: dns_ok, tls_ok, headers_ok.

util_security_txt_inspect(url, timeout)

Find security.txt contacts, disclosure policy, and trust links for a domain. Required: url.

util_seeded_ab_assignment(seed, subject, weight_a)

Seeded Ab Assignment assigns a stable subject to A or B using caller weights from bounded caller-supplied values without an external provider. Call Seeded Ab Assignment when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Ab Assignment as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: subject, weight_a, seed.

util_seeded_backoff_schedule(seed, factor, base_ms, attempts, jitter_fraction)

Seeded Backoff Schedule builds a reproducible jittered exponential backoff schedule from bounded caller-supplied values without an external provider. Call Seeded Backoff Schedule when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Backoff Schedule as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: base_ms, factor, attempts, jitter_fraction, seed.

util_seeded_balanced_groups(seed, items, group_count)

Seeded Balanced Groups distributes items reproducibly across near-equal groups from bounded caller-supplied values without an external provider. Call Seeded Balanced Groups when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Balanced Groups as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: items, group_count, seed.

util_seeded_batch_order(seed, batch_ids)

Seeded Batch Order orders batch identifiers reproducibly for replayable processing from bounded caller-supplied values without an external provider. Call Seeded Batch Order when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Batch Order as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: batch_ids, seed.

util_seeded_bootstrap_sample(seed, item_count, sample_count)

Seeded Bootstrap Sample samples indices with replacement for a reproducible bootstrap set from bounded caller-supplied values without an external provider. Call Seeded Bootstrap Sample when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Bootstrap Sample as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: item_count, sample_count, seed.

util_seeded_canary_sequence(seed, subjects, stage_sizes)

Seeded Canary Sequence orders subjects reproducibly for a staged canary rollout from bounded caller-supplied values without an external provider. Call Seeded Canary Sequence when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Canary Sequence as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: subjects, stage_sizes, seed.

util_seeded_candidate_subset(seed, count, candidates)

Seeded Candidate Subset selects a reproducible candidate subset while preserving original order from bounded caller-supplied values without an external provider. Call Seeded Candidate Subset when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Candidate Subset as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: candidates, count, seed.

util_seeded_coin_flip(seed, flips)

Seeded Coin Flip generates a reproducible sequence of binary outcomes from bounded caller-supplied values without an external provider. Call Seeded Coin Flip when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Coin Flip as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: flips, seed.

util_seeded_color_hex(seed)

Seeded Color Hex selects a reproducible display color without claiming accessibility from bounded caller-supplied values without an external provider. Call Seeded Color Hex when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Color Hex as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: seed.

util_seeded_consistent_shard(key, seed, shard_count)

Seeded Consistent Shard maps a stable key to one shard without mutable state from bounded caller-supplied values without an external provider. Call Seeded Consistent Shard when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Consistent Shard as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: key, shard_count, seed.

util_seeded_derangement(seed, items)

Seeded Derangement finds a reproducible permutation where no item keeps its position from bounded caller-supplied values without an external provider. Call Seeded Derangement when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Derangement as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: items, seed.

util_seeded_dice_roll(dice, seed, sides)

Seeded Dice Roll rolls bounded virtual dice reproducibly for simulations from bounded caller-supplied values without an external provider. Call Seeded Dice Roll when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Dice Roll as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: dice, sides, seed.

util_seeded_feature_bucket(seed, subject, bucket_count)

Seeded Feature Bucket maps a stable subject to an integer feature bucket from bounded caller-supplied values without an external provider. Call Seeded Feature Bucket when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Feature Bucket as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: subject, bucket_count, seed.

util_seeded_grid_point(seed, width, height)

Seeded Grid Point selects a reproducible integer coordinate inside rectangular bounds from bounded caller-supplied values without an external provider. Call Seeded Grid Point when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Grid Point as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: width, height, seed.

util_seeded_holdout_assignment(seed, subject, holdout_fraction)

Seeded Holdout Assignment assigns a stable subject to holdout or treatment from bounded caller-supplied values without an external provider. Call Seeded Holdout Assignment when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Holdout Assignment as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: subject, holdout_fraction, seed.

util_seeded_jitter_ms(seed, base_ms, jitter_fraction)

Seeded Jitter Ms adds reproducible bounded jitter to a base delay from bounded caller-supplied values without an external provider. Call Seeded Jitter Ms when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Jitter Ms as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: base_ms, jitter_fraction, seed.

util_seeded_kfold_split(seed, folds, item_count)

Seeded Kfold Split assigns item indices to reproducible cross-validation folds from bounded caller-supplied values without an external provider. Call Seeded Kfold Split when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Kfold Split as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: item_count, folds, seed.

util_seeded_load_spread(jobs, seed, workers)

Seeded Load Spread assigns jobs reproducibly across workers with near-even counts from bounded caller-supplied values without an external provider. Call Seeded Load Spread when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Load Spread as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: jobs, workers, seed.

util_seeded_lottery_draw(seed, labels, draw_count)

Seeded Lottery Draw draws unique labels reproducibly from a caller-owned pool without monetary claims from bounded caller-supplied values without an external provider. Call Seeded Lottery Draw when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Lottery Draw as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: labels, draw_count, seed.

util_seeded_matrix_cell(rows, seed, columns)

Seeded Matrix Cell selects one reproducible row and column inside matrix bounds from bounded caller-supplied values without an external provider. Call Seeded Matrix Cell when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Matrix Cell as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: rows, columns, seed.

util_seeded_multivariate_assignment(seed, subject, weights, variants)

Seeded Multivariate Assignment assigns a stable subject to one weighted variant from bounded caller-supplied values without an external provider. Call Seeded Multivariate Assignment when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Multivariate Assignment as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: subject, variants, weights, seed.

util_seeded_nonce_decimal(seed, maximum, minimum)

Seeded Nonce Decimal derives a reproducible non-security decimal nonce inside bounds from bounded caller-supplied values without an external provider. Call Seeded Nonce Decimal when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Nonce Decimal as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: minimum, maximum, seed.

util_seeded_pairing(seed, participants)

Seeded Pairing creates reproducible non-overlapping pairs and reports an optional bye from bounded caller-supplied values without an external provider. Call Seeded Pairing when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Pairing as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: participants, seed.

util_seeded_partition(seed, items, partitions)

Seeded Partition assigns each caller item reproducibly into a named partition from bounded caller-supplied values without an external provider. Call Seeded Partition when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Partition as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: items, partitions, seed.

util_seeded_passphrase_words(seed, count, words)

Seeded Passphrase Words selects reproducible words from a caller list for non-secret labels from bounded caller-supplied values without an external provider. Call Seeded Passphrase Words when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Passphrase Words as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: words, count, seed.

util_seeded_percentage_rollout(seed, subject, percentage)

Seeded Percentage Rollout evaluates stable inclusion in a percentage rollout from bounded caller-supplied values without an external provider. Call Seeded Percentage Rollout when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Percentage Rollout as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: subject, percentage, seed.

util_seeded_permutation_index(seed, permutation_count)

Seeded Permutation Index selects one reproducible permutation index within a caller count from bounded caller-supplied values without an external provider. Call Seeded Permutation Index when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Permutation Index as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: permutation_count, seed.

util_seeded_priority_order(seed, tasks)

Seeded Priority Order orders equal-priority tasks reproducibly from bounded caller-supplied values without an external provider. Call Seeded Priority Order when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Priority Order as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: tasks, seed.

util_seeded_queue_pick(seed, queues)

Seeded Queue Pick selects one queue reproducibly from available queue names from bounded caller-supplied values without an external provider. Call Seeded Queue Pick when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Queue Pick as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: queues, seed.

util_seeded_random_bytes_hex(seed, byte_count)

Seeded Random Bytes Hex returns reproducible non-cryptographic bytes encoded as hex from bounded caller-supplied values without an external provider. Call Seeded Random Bytes Hex when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Random Bytes Hex as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: byte_count, seed.

util_seeded_random_float(seed, count, maximum, minimum)

Seeded Random Float returns reproducible floats inside caller bounds from bounded caller-supplied values without an external provider. Call Seeded Random Float when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Random Float as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: minimum, maximum, count, seed.

util_seeded_random_sample(seed, count, items)

Select a reproducible sample without replacement from a bounded caller-supplied list. Required: items, count, seed.

util_seeded_range_sequence(seed, count, maximum, minimum)

Seeded Range Sequence generates unique reproducible integers from a bounded range from bounded caller-supplied values without an external provider. Call Seeded Range Sequence when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Range Sequence as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: minimum, maximum, count, seed.

util_seeded_rendezvous_rank(key, seed, nodes)

Seeded Rendezvous Rank ranks nodes for a key using reproducible rendezvous hashing from bounded caller-supplied values without an external provider. Call Seeded Rendezvous Rank when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Rendezvous Rank as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: key, nodes, seed.

util_seeded_reservoir_sample(seed, count, items)

Seeded Reservoir Sample selects a bounded reproducible sample from caller items from bounded caller-supplied values without an external provider. Call Seeded Reservoir Sample when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Reservoir Sample as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: items, count, seed.

util_seeded_round_robin_start(seed, member_count)

Seeded Round Robin Start chooses a reproducible starting index for round-robin rotation from bounded caller-supplied values without an external provider. Call Seeded Round Robin Start when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Round Robin Start as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: member_count, seed.

util_seeded_sample_indices(seed, count, population_size)

Seeded Sample Indices samples unique zero-based indices from a population size from bounded caller-supplied values without an external provider. Call Seeded Sample Indices when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Sample Indices as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: population_size, count, seed.

util_seeded_sample_with_replacement(seed, count, items)

Seeded Sample With Replacement samples caller items reproducibly with replacement from bounded caller-supplied values without an external provider. Call Seeded Sample With Replacement when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Sample With Replacement as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: items, count, seed.

util_seeded_schedule_offset(seed, window_seconds)

Seeded Schedule Offset selects a reproducible offset inside a scheduling window from bounded caller-supplied values without an external provider. Call Seeded Schedule Offset when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Schedule Offset as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: window_seconds, seed.

util_seeded_shuffle(seed, items)

Seeded Shuffle produces a reproducible permutation of caller items from bounded caller-supplied values without an external provider. Call Seeded Shuffle when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Shuffle as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: items, seed.

util_seeded_stratified_split(seed, labels, test_fraction)

Seeded Stratified Split splits labeled items while preserving per-label proportions from bounded caller-supplied values without an external provider. Call Seeded Stratified Split when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Stratified Split as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: labels, test_fraction, seed.

util_seeded_tie_break(seed, candidates)

Seeded Tie Break breaks a tie reproducibly among equally scored candidates from bounded caller-supplied values without an external provider. Call Seeded Tie Break when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Tie Break as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: candidates, seed.

util_seeded_time_slot_pick(seed, slots)

Seeded Time Slot Pick selects a reproducible slot from caller-provided time labels from bounded caller-supplied values without an external provider. Call Seeded Time Slot Pick when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Time Slot Pick as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: slots, seed.

util_seeded_token(seed, length, alphabet)

Seeded Token builds a reproducible non-secret token from a caller alphabet from bounded caller-supplied values without an external provider. Call Seeded Token when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Token as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: alphabet, length, seed.

util_seeded_train_test_split(seed, items, test_fraction)

Seeded Train Test Split splits items into reproducible train and test sets from bounded caller-supplied values without an external provider. Call Seeded Train Test Split when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Train Test Split as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: items, test_fraction, seed.

util_seeded_uuid(name, seed, namespace)

Seeded UUID derives a reproducible UUID-shaped identifier from seed and namespace text from bounded caller-supplied values without an external provider. Call Seeded UUID when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded UUID as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: namespace, name, seed.

util_seeded_weighted_choice(seed, items, weights)

Seeded Weighted Choice selects one item using non-negative caller weights from bounded caller-supplied values without an external provider. Call Seeded Weighted Choice when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Weighted Choice as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: items, weights, seed.

util_seeded_weighted_partition(seed, items, weights, partitions)

Seeded Weighted Partition assigns items to partitions using reproducible capacity weights from bounded caller-supplied values without an external provider. Call Seeded Weighted Partition when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Weighted Partition as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: items, partitions, weights, seed.

util_seeded_weighted_sample(seed, count, items, weights)

Seeded Weighted Sample samples unique items using caller weights without replacement from bounded caller-supplied values without an external provider. Call Seeded Weighted Sample when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Weighted Sample as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: items, weights, count, seed.

util_seeded_window_pick(seed, items, window_size)

Seeded Window Pick selects a reproducible contiguous window from caller items from bounded caller-supplied values without an external provider. Call Seeded Window Pick when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Window Pick as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: items, window_size, seed.

util_seeded_winners_without_replacement(seed, candidates, winner_count)

Seeded Winners Without Replacement draws an ordered reproducible winner list without replacement from bounded caller-supplied values without an external provider. Call Seeded Winners Without Replacement when an agent needs a reproducible decision that another run can independently replay. Returns the deterministic selection plus seed provenance, population bounds, and the operation-specific outcome for Seeded Winners Without Replacement as versioned deterministic JSON. Limitation: Uses Python pseudorandomness for reproducibility, not cryptography, lotteries with value, security tokens, gambling, or custody. Required: candidates, winner_count, seed.

util_selector_extract(calldata)

Extract the 4-byte function selector from calldata hex. Required: calldata.

util_semver_bump(changes)

Recommend a semantic-version bump from explicit change signals. Required: changes.

util_semver_compare(a, b)

Compare two semantic versions (major.minor.patch). Required: a, b.

util_semver_compare_lite(a, b)

Compare two semver triples (-1,0,1) Required: a, b.

util_semver_parts(version)

Parse major.minor.patch from a semver string Required: version.

util_sentence_case(text)

Delx first-party micro-utility util_sentence_case at $0.001 USDC. Required: text.

util_server_timing_parse(header)

Server Timing Parse parses bounded Server-Timing metrics and durations from bounded caller-supplied values without an external provider. Call Server Timing Parse before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Server Timing Parse as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header.

util_set_diff(a, b)

Set difference a - b Required: a, b.

util_set_intersect(a, b)

Set intersection of two lists Required: a, b.

util_set_union(a, b)

Set union of two lists (order-preserving) Required: a, b.

util_sha1_hex(text)

SHA-1 hex digest of text Required: text.

util_sha256_hex(text)

SHA-256 hex digest of text Required: text.

util_shares_from_budget(price, budget)

Shares purchasable for a budget at a given share price Required: budget, price.

util_sharpe_lite(returns)

Lite Sharpe = mean/stdev of a return sample (no rf, sample) Required: returns.

util_sigmoid(x)

Sigmoid 1/(1+e^{−x}) back to probability space Required: x.

util_sign_n(value)

Sign of a number (-1, 0, 1) Required: value.

util_signal_threshold(lower, upper, value)

Classify a signal against optional upper and lower thresholds. Required: value.

util_similarity

Normalized string similarity (Levenshtein). Required: a, b.

util_sitemap_changefreq_check(changefreq)

Sitemap Changefreq Check classifies a sitemap changefreq token against the defined vocabulary from bounded caller-supplied values without an external provider. Call Sitemap Changefreq Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Sitemap Changefreq Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: changefreq.

util_sitemap_entry_validate(entry)

Sitemap Entry Validate checks one sitemap URL entry for required location and bounded metadata from bounded caller-supplied values without an external provider. Call Sitemap Entry Validate before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Sitemap Entry Validate as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: entry.

util_sitemap_lastmod_freshness(now_epoch, lastmod_epoch, maximum_age_seconds)

Sitemap Lastmod Freshness calculates sitemap lastmod age against a caller freshness budget from bounded caller-supplied values without an external provider. Call Sitemap Lastmod Freshness before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Sitemap Lastmod Freshness as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: lastmod_epoch, now_epoch, maximum_age_seconds.

util_sitemap_probe(url, timeout)

Check sitemap and crawl-structure hints fast to see how a site exposes crawlable structure. Required: url.

util_skill_md_stub(title, tool_names)

Generate a minimal skill.md body for a product cluster Required: title, tool_names.

util_sla_breach_count(sla_ms, samples_ms)

Count SLA breaches in a latency sample against a millisecond threshold. Required: samples_ms, sla_ms.

util_sla_budget_remaining(now_ts, deadline_ts)

Remaining SLA seconds from deadline and now Required: deadline_ts, now_ts.

util_slippage_bps(mid, fill)

Slippage in basis points versus mid for execution reports Required: mid, fill.

util_slippage_pct(mid, fill)

Slippage percent of expected mid versus fill/impact price Required: mid, fill.

util_slo_compliance(good_events, total_events, target_percent)

Measure achieved reliability against an SLO target from good and total events. Required: good_events, total_events.

util_slug_from_url(url)

Delx first-party micro-utility util_slug_from_url at $0.001 USDC. Required: url.

util_slug_validate(slug, max_len)

Validate a URL/product slug against a safe lowercase hyphen pattern Required: slug.

util_slugify(text, lowercase, separator, max_length)

Convert text to a URL-safe slug. Required: text.

util_slugify_lite(text)

Slugify text to lowercase hyphen form Required: text.

util_snake_to_camel(text)

Convert snake_case identifiers to lowerCamelCase Required: text.

util_sort_keys(object)

Return object keys sorted lexicographically for stable hashes Required: object.

util_sort_lines

Sort lines of text. Required: text.

util_sortino_lite(returns)

Lite Sortino using downside stdev of negative returns Required: returns.

util_soundex(text)

Delx first-party micro-utility util_soundex at $0.001 USDC. Required: text.

util_spf_record_flags(spf)

Detect common SPF mechanisms present in a TXT record Required: spf.

util_split_lines(text)

Split text into lines Required: text.

util_spread_abs(ask, bid)

Absolute bid–ask spread for liquidity snapshots Required: bid, ask.

util_spread_bps(ask, bid)

Compute mid price and bid-ask spread in basis points. Required: bid, ask.

util_spread_pct(ask, bid)

Bid–ask spread as percent of mid for cross-asset compare Required: bid, ask.

util_sql_injection_scan(text)

Screen untrusted text for common SQL-injection signals without echoing it. Required: text.

util_ssrf_url_audit(url)

Screen a URL for private hosts, unsafe schemes, and embedded credentials. Required: url.

util_stable_hash(text)

Canonical SHA-256 of JSON-stable payload Required: text.

util_starts_with(text, prefix)

Check whether text starts with a prefix Required: text, prefix.

util_status_class(status_code)

Classify an HTTP status code into family with retry hint Required: status_code.

util_status_retry_classify(method, status, retry_statuses)

Status Retry Classify classifies whether a method and status pair is retryable under caller policy from bounded caller-supplied values without an external provider. Call Status Retry Classify before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Status Retry Classify as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: method, status, retry_statuses.

util_stdev(values)

Delx first-party micro-utility util_stdev at $0.001 USDC. Required: values.

util_step_budget_plan(steps, weights, total_budget_ms)

Allocate a total millisecond budget across ordered workflow steps using optional weights. Required: steps, total_budget_ms.

util_stop_loss_pct(pct, side, entry)

Stop-loss price from entry and stop percent move Required: entry, pct, side.

util_string_pad(fill, side, text, width)

Pad a string to a target width with a fill character Required: text, width.

util_string_trunc(text, ellipsis, max_chars)

Truncate a string to max_chars with optional ellipsis Required: text.

util_strip_accents(text)

Delx first-party micro-utility util_strip_accents at $0.001 USDC. Required: text.

util_strip_html_lite(text)

Strip HTML tags from caller-held markup Required: text.

util_structured_log_summary(log_text)

Summarize log severity and hashed error fingerprints without returning sensitive error lines. Required: log_text.

util_success_rate(total, success)

Compute success and failure rates from success and total counts. Required: success, total.

util_sum_n(values)

Sum of a numeric list Required: values.

util_swap_case(text)

Swap letter case for deterministic transform tests Required: text.

util_take_profit_pct(pct, side, entry)

Take-profit price from entry and target percent move Required: entry, pct, side.

util_task_priority_score(effort, impact, urgency)

Score task priority from urgency, impact, and effort into an ordered band. Required: urgency, impact, effort.

util_temp_c_to_f(celsius)

Convert Celsius to Fahrenheit Required: celsius.

util_temp_f_to_c(fahrenheit)

Convert Fahrenheit to Celsius Required: fahrenheit.

util_text_chunk_plan(text, overlap, chunk_size)

Plan overlapping character chunks for RAG without storing the text. Required: text, chunk_size.

util_text_reverse(text)

Reverse a text string for deterministic transform tests Required: text.

util_text_sha256(text)

SHA-256 hex digest of UTF-8 text Required: text.

util_text_truncate

Truncate text to a max length with optional ellipsis. Required: text, max_length.

util_ticker_canonicalize(ticker)

Canonicalize a ticker symbol and split common base/quote separators. Required: ticker.

util_time_series_gap_scan(timestamps, expected_seconds)

Detect gaps larger than an expected interval in ISO-8601 timestamps. Required: timestamps, expected_seconds.

util_timeout_budget(steps, total_ms, reserved_ms)

Split a total timeout into per-step budgets after reserved overhead. Required: total_ms, steps.

util_timestamp_align(timestamps, interval_seconds)

Align ISO-8601 timestamps down to fixed UTC interval buckets. Required: timestamps, interval_seconds.

util_timestamp_convert(to, input)

Convert between timestamp formats: Unix epoch, ISO 8601, and human-readable. Required: input.

util_timezone_lookup(at, timezone)

Resolve an IANA timezone to offset, abbreviation, DST flag, and local time for a given instant (or now). Required: timezone.

util_title_case(text)

Delx first-party micro-utility util_title_case at $0.001 USDC. Required: text.

util_title_words(text)

Title-case words for headings and display labels Required: text.

util_tls_inspect(url, timeout)

Inspect TLS issuer, subject, SANs, and expiry to check trust and renewal risk. Required: url.

util_token_amount_to_raw(amount, decimals)

Scale human token amounts by decimals into on-chain raw units Required: amount, decimals.

util_token_budget_split(total_tokens, system_tokens, reserved_output_tokens)

Split a model token budget into system, output reserve, and available input. Required: total_tokens.

util_token_echo_1(text)

Echo text with length (probe channel 1) Required: text.

util_token_echo_2(text)

Echo text with length (probe channel 2) Required: text.

util_token_estimate(text, model)

Estimate token count for text. Uses word/4 heuristic (GPT-family) and char/4 (Claude-family). Required: text.

util_token_raw_to_amount(raw, decimals)

Scale raw token units by decimals into human-readable amounts Required: raw, decimals.

util_tool_permission_audit(allowed, requested)

Compare requested agent tools with an explicit least-privilege allowlist. Required: requested, allowed.

util_transfer_encoding_check(header)

Transfer Encoding Check checks transfer-coding order and whether chunked appears only last from bounded caller-supplied values without an external provider. Call Transfer Encoding Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Transfer Encoding Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header.

util_trim_collapse(text)

Trim ends and collapse internal whitespace runs to single spaces Required: text.

util_truncate_middle(text, max_chars)

Truncate long text in the middle with an ellipsis Required: text, max_chars.

util_tsv_row(fields)

Join fields into a TSV row Required: fields.

util_twap_simple(prices)

Simple average of price samples for a light TWAP mark Required: prices.

util_tx_hash_format_check(tx_hash)

Validate 0x + 64 hex transaction hash shape Required: tx_hash.

util_tx_hash_validate(tx_hash)

Validate a 32-byte transaction hash format and normalize casing. Required: tx_hash.

util_tx_value_eth(wei)

Convert a transaction value wei field into ether for display Required: wei.

util_tz_offset_fixed(offset_minutes)

Format a fixed UTC offset as ±HH:MM Required: offset_minutes.

util_ulid_generate(count)

Generate one or more Crockford Base32 ULIDs.

util_unique_sorted(items)

Unique and sort a JSON list Required: items.

util_unique_words(text)

Delx first-party micro-utility util_unique_words at $0.001 USDC. Required: text.

util_unit_convert(value, to_unit, from_unit, conversions)

Convert values between units (length, mass, volume, time, data, temperature). Supports one conversion or a batch of up to 100.

util_unit_economics(cost, price)

Calculate gross profit and margin from per-unit price and cost. Required: price, cost.

util_unix_now

Delx first-party micro-utility util_unix_now at $0.001 USDC.

util_unix_to_iso(unix)

Convert unix seconds to ISO-8601 Z Required: unix.

util_unrealized_pnl(mark, size, entry)

Unrealized PnL versus mark from entry and size Required: entry, mark, size.

util_unslugify_lite(text)

Turn slugs into Title Case words Required: text.

util_url_allowlist_match(url, allowlist)

Match a URL against host/path allowlist patterns before egress Required: url, allowlist.

util_url_canonical_compare(left, right)

URL Canonical Compare compares URLs after scheme, host, port, path, and query normalization from bounded caller-supplied values without an external provider. Call URL Canonical Compare before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for URL Canonical Compare as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: left, right.

util_url_decode_safe(text)

Percent-decode text Required: text.

util_url_encode

URL-encode or URL-decode text. Required: text.

util_url_encode_component(text)

Delx first-party micro-utility util_url_encode_component at $0.001 USDC. Required: text.

util_url_encode_safe(text)

Percent-encode text for query components Required: text.

util_url_health(url, timeout)

Check if a URL is reachable. Returns HTTP status, latency, and key headers. Required: url.

util_url_join(base, path)

Join a base URL with a relative path or href. Required: base.

util_url_origin(url)

Extract scheme://host[:port] origin from a URL without network access Required: url.

util_url_parse(url)

Delx first-party micro-utility util_url_parse at $0.001 USDC. Required: url.

util_url_parts(url)

Split a URL into scheme/netloc/path/query/fragment Required: url.

util_url_path_normalize(url)

Normalize URL scheme, host casing, and path slashes without fetching. Required: url.

util_url_port_policy_check(url, allowed_ports)

URL Port Policy Check checks a URL effective port against a caller allowlist from bounded caller-supplied values without an external provider. Call URL Port Policy Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for URL Port Policy Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: url, allowed_ports.

util_url_query_param_diff(after_url, before_url)

URL Query Param Diff finds added, removed, and changed query parameters between URLs from bounded caller-supplied values without an external provider. Call URL Query Param Diff before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for URL Query Param Diff as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: before_url, after_url.

util_url_redirect_target_check(target, base_url, allowed_hosts, allowed_schemes)

URL Redirect Target Check checks a redirect target against allowed schemes and hosts from bounded caller-supplied values without an external provider. Call URL Redirect Target Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for URL Redirect Target Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: base_url, target, allowed_hosts, allowed_schemes.

util_url_tracking_param_strip(url, parameters)

URL Tracking Param Strip removes caller-selected tracking parameters while preserving the rest from bounded caller-supplied values without an external provider. Call URL Tracking Param Strip before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for URL Tracking Param Strip as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: url, parameters.

util_us_market_holiday_check(date)

Check if a YYYY-MM-DD is in a static US market holiday set Required: date.

util_usdc_raw_to_amount(raw)

Convert 6-decimal USDC raw units into human USDC amounts Required: raw.

util_usdc_to_raw(usdc)

Convert USDC human amounts into 6-decimal raw units used on-chain Required: usdc.

util_usgs_quake_stub_parse(properties)

Parse magnitude/place from a USGS-like feature properties object Required: properties.

util_uuid_generate(count)

Generate one or more UUIDv4 strings.

util_values_types(object)

Map object keys to JSON type names Required: object.

util_vary_header_normalize(header)

Vary Header Normalize normalizes and deduplicates Vary header field names from bounded caller-supplied values without an external provider. Call Vary Header Normalize before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Vary Header Normalize as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header.

util_vesting_remaining(total, end_ts, now_ts, cliff_ts, start_ts)

Remaining unlock amount for linear vesting with optional cliff Required: total, start_ts, end_ts, now_ts.

util_vin_decode(vin)

Decode a 17-character VIN via NHTSA vPIC into make, model, year, body class, and related fields with attribution. Required: vin.

util_volatility_sample(returns, periods_per_year)

Estimate descriptive sample volatility from caller-supplied returns. Required: returns.

util_vwap_simple(samples)

Volume-weighted average price from price/size samples Required: samples.

util_webhook_delivery_plan(event)

Produce a signature, idempotency, acknowledgement, and retry plan for webhooks. Required: event.

util_webhook_event_normalize(event)

Normalize heterogeneous webhook events and derive a stable payload fingerprint. Required: event.

util_webhook_signature_timestamp_check(now_epoch, timestamp_epoch, maximum_age_seconds, maximum_future_skew_seconds)

Webhook Signature Timestamp Check checks webhook timestamp age and future skew before signature verification from bounded caller-supplied values without an external provider. Call Webhook Signature Timestamp Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Webhook Signature Timestamp Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: timestamp_epoch, now_epoch, maximum_age_seconds, maximum_future_skew_seconds.

util_website_intelligence_report(url, timeout)

Composite website intelligence report with page, social, link, form, feed, and contact signals. Required: url.

util_weekday_name(iso)

Weekday name from ISO date Required: iso.

util_wei_ether_convert(amount, direction)

Convert between wei and ether using fixed 1e18 scaling without RPC. Required: amount.

util_wei_to_eth(wei)

Scale wei integer strings into ether for dashboards and receipts Required: wei.

util_wei_to_gwei(wei)

Convert wei integer strings into gwei for readable gas quotes Required: wei.

util_weight_normalize(weights)

Normalize non-negative weights so they sum to one when possible. Required: weights.

util_weighted_choice(items, weights)

Pick one item from a list using non-negative weights Required: items.

util_win_rate(pnls)

Win rate from wins and total trades Required: pnls.

util_wmi_decode(wmi)

Decode a 3-character WMI (World Manufacturer Identifier) via NHTSA vPIC into manufacturer and vehicle type. Required: wmi.

util_word_count(text)

Count words, characters, lines and paragraphs in text. Required: text.

util_word_frequency(text)

Delx first-party micro-utility util_word_frequency at $0.001 USDC. Required: text.

util_workflow_gate(checks, required)

Gate a workflow on an explicit boolean check map before side effects. Required: checks.

util_wrap_text(text)

Delx first-party micro-utility util_wrap_text at $0.001 USDC. Required: text.

util_www_authenticate_parse(header)

Www Authenticate Parse parses an authentication challenge scheme and bounded parameters from bounded caller-supplied values without an external provider. Call Www Authenticate Parse before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for Www Authenticate Parse as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header.

util_x_content_type_options_check(value)

X Content Type Options Check checks whether X-Content-Type-Options is exactly nosniff from bounded caller-supplied values without an external provider. Call X Content Type Options Check before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for X Content Type Options Check as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: value.

util_x_forwarded_for_chain(header)

X Forwarded For Chain normalizes a proxy chain without deciding whether an address is trusted from bounded caller-supplied values without an external provider. Call X Forwarded For Chain before accepting, caching, redirecting, or retrying a caller-supplied web response. Returns normalized web evidence, the computed finding, and an explicit pass or advisory status for X Forwarded For Chain as versioned deterministic JSON. Limitation: Evaluates only caller-supplied metadata; it performs no DNS lookup, HTTP request, web search, credential use, or persistent storage. Required: header.

util_x402_accepts_rank(accepts, preferred_assets, preferred_networks)

X402 Accepts Rank ranks challenge accepts entries against buyer network and asset preferences from bounded caller-supplied values without an external provider. Call X402 Accepts Rank before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Accepts Rank as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: accepts, preferred_networks, preferred_assets.

util_x402_amount_budget_check(amount_raw, maximum_raw)

X402 Amount Budget Check checks challenge raw amount against an exact buyer call budget from bounded caller-supplied values without an external provider. Call X402 Amount Budget Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Amount Budget Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: amount_raw, maximum_raw.

util_x402_asset_match(asset, supported_assets)

X402 Asset Match compares a challenge asset identifier with buyer-supported assets from bounded caller-supplied values without an external provider. Call X402 Asset Match before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Asset Match as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: asset, supported_assets.

util_x402_catalog_route_check(route)

X402 Catalog Route Check checks one catalog route for tool, method, path, and positive price from bounded caller-supplied values without an external provider. Call X402 Catalog Route Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Catalog Route Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: route.

util_x402_challenge_shape_check(challenge)

X402 Challenge Shape Check checks required top-level fields in a caller-supplied payment challenge from bounded caller-supplied values without an external provider. Call X402 Challenge Shape Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Challenge Shape Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: challenge.

util_x402_challenge_version_check(version, supported_versions)

X402 Challenge Version Check compares challenge protocol version with supported versions from bounded caller-supplied values without an external provider. Call X402 Challenge Version Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Challenge Version Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: version, supported_versions.

util_x402_copy_completeness_check(description, service_name, minimum_length)

X402 Copy Completeness Check checks product copy for service name, job, outcome, limitation, and minimum length from bounded caller-supplied values without an external provider. Call X402 Copy Completeness Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Copy Completeness Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: service_name, description, minimum_length.

util_x402_delivery_content_type_check(accepted, content_type)

X402 Delivery Content Type Check compares delivery content type with buyer-accepted media types from bounded caller-supplied values without an external provider. Call X402 Delivery Content Type Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Delivery Content Type Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: content_type, accepted.

util_x402_delivery_hash_check(text, expected_sha256)

X402 Delivery Hash Check verifies delivered text against a caller-supplied sha-256 digest from bounded caller-supplied values without an external provider. Call X402 Delivery Hash Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Delivery Hash Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: text, expected_sha256.

util_x402_delivery_latency_check(latency_ms, maximum_ms)

X402 Delivery Latency Check checks observed delivery latency against a buyer SLA from bounded caller-supplied values without an external provider. Call X402 Delivery Latency Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Delivery Latency Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: latency_ms, maximum_ms.

util_x402_delivery_receipt_check(receipt)

X402 Delivery Receipt Check checks receipt fields for payment id, resource, amount, and delivery status from bounded caller-supplied values without an external provider. Call X402 Delivery Receipt Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Delivery Receipt Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: receipt.

util_x402_delivery_schema_check(result, required_fields)

X402 Delivery Schema Check checks delivered object fields against a caller-required field list from bounded caller-supplied values without an external provider. Call X402 Delivery Schema Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Delivery Schema Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: result, required_fields.

util_x402_delivery_size_check(result, maximum_bytes)

X402 Delivery Size Check checks serialized result byte size against a buyer limit from bounded caller-supplied values without an external provider. Call X402 Delivery Size Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Delivery Size Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: result, maximum_bytes.

util_x402_delivery_status_check(status, accepted_statuses)

X402 Delivery Status Check classifies delivery HTTP status against accepted success codes from bounded caller-supplied values without an external provider. Call X402 Delivery Status Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Delivery Status Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: status, accepted_statuses.

util_x402_discovery_parity_check(catalog_tools, openapi_tools, runtime_tools, portfolio_tools)

X402 Discovery Parity Check compares tool-name sets from catalog, OpenAPI, portfolio, and runtime from bounded caller-supplied values without an external provider. Call X402 Discovery Parity Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Discovery Parity Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: catalog_tools, openapi_tools, portfolio_tools, runtime_tools.

util_x402_idempotency_check(payment_key, delivery_key)

X402 Idempotency Check compares payment and delivery idempotency keys for one operation from bounded caller-supplied values without an external provider. Call X402 Idempotency Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Idempotency Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: payment_key, delivery_key.

util_x402_network_match(network, supported_networks)

X402 Network Match compares a challenge network identifier with buyer-supported networks from bounded caller-supplied values without an external provider. Call X402 Network Match before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Network Match as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: network, supported_networks.

util_x402_openapi_route_check(operation)

X402 Openapi Route Check checks one OpenAPI operation for tool id, request schema, and 402 response from bounded caller-supplied values without an external provider. Call X402 Openapi Route Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Openapi Route Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: operation.

util_x402_pay_to_match(pay_to, expected)

X402 Pay To Match compares challenge pay-to address with an expected merchant address from bounded caller-supplied values without an external provider. Call X402 Pay To Match before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Pay To Match as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: pay_to, expected.

util_x402_payment_deadline_check(now_epoch, deadline_epoch, maximum_lead_seconds, minimum_lead_seconds)

X402 Payment Deadline Check checks payment authorization deadline against caller time budgets from bounded caller-supplied values without an external provider. Call X402 Payment Deadline Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Payment Deadline Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: deadline_epoch, now_epoch, minimum_lead_seconds, maximum_lead_seconds.

util_x402_payment_header_shape(header)

X402 Payment Header Shape checks a caller-supplied payment header for bounded base64 or JSON shape from bounded caller-supplied values without an external provider. Call X402 Payment Header Shape before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Payment Header Shape as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: header.

util_x402_payment_nonce_check(nonce, seen_nonces, minimum_length)

X402 Payment Nonce Check checks a payment nonce for presence, length, and caller-known reuse from bounded caller-supplied values without an external provider. Call X402 Payment Nonce Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Payment Nonce Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: nonce, seen_nonces, minimum_length.

util_x402_payment_payload_size(payload, maximum_bytes)

X402 Payment Payload Size checks serialized payment payload bytes against a buyer limit from bounded caller-supplied values without an external provider. Call X402 Payment Payload Size before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Payment Payload Size as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: payload, maximum_bytes.

util_x402_payment_preflight(requirements, asset_decimals, max_amount_usdc, allowed_networks, now_epoch_seconds, min_expiry_seconds)

Evaluate caller-supplied x402 accepts against price, network, recipient, scheme, and optional expiry policy. Required: requirements, max_amount_usdc, allowed_networks.

util_x402_payment_replay_key(nonce, payer, resource)

X402 Payment Replay Key derives a stable non-secret replay-detection key from payment metadata from bounded caller-supplied values without an external provider. Call X402 Payment Replay Key before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Payment Replay Key as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: payer, nonce, resource.

util_x402_portfolio_route_check(product)

X402 Portfolio Route Check checks one portfolio entry for ownership, economics, resource, and discoverability from bounded caller-supplied values without an external provider. Call X402 Portfolio Route Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Portfolio Route Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: product.

util_x402_price_drift_check(advertised_raw, challenged_raw, maximum_drift_raw)

X402 Price Drift Check compares advertised and challenged raw prices against a tolerance from bounded caller-supplied values without an external provider. Call X402 Price Drift Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Price Drift Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: advertised_raw, challenged_raw, maximum_drift_raw.

util_x402_refund_amount_check(settled_raw, requested_raw)

X402 Refund Amount Check compares requested refund raw amount with original settled amount from bounded caller-supplied values without an external provider. Call X402 Refund Amount Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Refund Amount Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: requested_raw, settled_raw.

util_x402_refund_deadline_check(window_seconds, settled_at_epoch, requested_at_epoch)

X402 Refund Deadline Check checks a refund request timestamp against a policy window from bounded caller-supplied values without an external provider. Call X402 Refund Deadline Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Refund Deadline Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: settled_at_epoch, requested_at_epoch, window_seconds.

util_x402_refund_eligibility_check(settled, delivered, failure_code, eligible_codes)

X402 Refund Eligibility Check evaluates caller-supplied delivery outcome against explicit refund reasons from bounded caller-supplied values without an external provider. Call X402 Refund Eligibility Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Refund Eligibility Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: settled, delivered, failure_code, eligible_codes.

util_x402_refund_receipt_check(receipt)

X402 Refund Receipt Check checks caller-supplied refund receipt identifiers, amount, and status from bounded caller-supplied values without an external provider. Call X402 Refund Receipt Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Refund Receipt Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: receipt.

util_x402_registry_route_check(route)

X402 Registry Route Check checks one runtime registry entry for tool, path, method, and upstream from bounded caller-supplied values without an external provider. Call X402 Registry Route Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Registry Route Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: route.

util_x402_resource_match(expected, resource)

X402 Resource Match compares challenge resource URL with the intended paid resource from bounded caller-supplied values without an external provider. Call X402 Resource Match before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Resource Match as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: resource, expected.

util_x402_resource_summary(url, timeout)

Summarize a server's .well-known/x402 resources, pricing surface, networks, and paths. Required: url.

util_x402_response_header_shape(header)

X402 Response Header Shape checks a payment-response header for bounded base64 or JSON shape from bounded caller-supplied values without an external provider. Call X402 Response Header Shape before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Response Header Shape as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: header.

util_x402_retry_safety_check(payment_settled, delivery_received, idempotency_key_present)

X402 Retry Safety Check checks whether retrying after challenge or delivery failure risks double settlement from bounded caller-supplied values without an external provider. Call X402 Retry Safety Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Retry Safety Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: payment_settled, delivery_received, idempotency_key_present.

util_x402_scheme_select(schemes, supported_schemes)

X402 Scheme Select selects the first buyer-supported payment scheme from challenge options from bounded caller-supplied values without an external provider. Call X402 Scheme Select before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Scheme Select as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: schemes, supported_schemes.

util_x402_server_audit(url, timeout)

Audit an x402 server with discovery, pricing, reliability, and documentation readiness signals. Required: url.

util_x402_server_probe(url, timeout)

Probe an x402 server end-to-end: discovery, status, tools, reliability, and OpenAPI. Required: url.

util_x402_settlement_amount_check(settled_raw, accepted_raw)

X402 Settlement Amount Check compares settled raw amount with the accepted challenge amount from bounded caller-supplied values without an external provider. Call X402 Settlement Amount Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Settlement Amount Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: settled_raw, accepted_raw.

util_x402_settlement_network_check(accepted_network, settlement_network)

X402 Settlement Network Check compares settlement chain with the accepted challenge network from bounded caller-supplied values without an external provider. Call X402 Settlement Network Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Settlement Network Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: settlement_network, accepted_network.

util_x402_settlement_recipient_check(pay_to, recipient)

X402 Settlement Recipient Check compares settlement recipient with the accepted pay-to address from bounded caller-supplied values without an external provider. Call X402 Settlement Recipient Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Settlement Recipient Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: recipient, pay_to.

util_x402_settlement_tx_shape(network, transaction)

X402 Settlement Tx Shape checks a settlement transaction identifier for expected chain shape from bounded caller-supplied values without an external provider. Call X402 Settlement Tx Shape before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Settlement Tx Shape as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: network, transaction.

util_x402_timeout_check(now_epoch, expires_epoch, minimum_lead_seconds)

X402 Timeout Check checks challenge expiration and minimum buyer processing time from bounded caller-supplied values without an external provider. Call X402 Timeout Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for X402 Timeout Check as versioned deterministic JSON. Limitation: Validates metadata only; it never settles payment, signs payloads, contacts a facilitator, refunds funds, or asserts on-chain truth. Required: expires_epoch, now_epoch, minimum_lead_seconds.

util_xml_escape(text)

XML-escape text for safe embedding Required: text.

util_xss_signal_scan(text)

Screen untrusted text for common cross-site scripting signals without echoing it. Required: text.

util_zip_keys_values(keys, values)

Zip key and value lists into an object Required: keys, values.

wellness_webhook(events, threshold, session_id, callback_url, cooldown_min, entropy_threshold)

Subscribe to proactive reliability alerts to reduce polling overhead. Pricing is dynamic; check /api/v1/tools. Required: session_id, callback_url.

Last successful function declaration observed on 2026-08-13. Source: https://api.delx.ai/v1/mcp. We list what the server declared; we do not call any of these functions.

Endpoint status observed on 2026-08-13. Source: https://api.delx.ai/v1/mcp.

Signals

These are separate measurements of different things. They are deliberately not combined into one score, because a popularity number that mixes website traffic with saves and stars cannot be checked or acted on.

Where to get it

• Repository — https://github.com/davidmosiah/delx-protocol
• Website — https://commerce.delx.ai/
• Remote endpoint (streamable-http) — https://api.delx.ai/v1/mcp

Also from davidmosiah

• io.github.davidmosiah/agent-seo-engine — last commit 2026-07-03
  Agent-first local SEO quality, intent and opportunity engine with CLI and optional MCP server.
• io.github.davidmosiah/apple-health-mcp — last commit 2026-08-05
  Unofficial local-first MCP server for Apple Health export data.
• io.github.davidmosiah/astral-mcp — last commit 2026-08-05
  Precision-audited astrology MCP: natal charts, transits, synastry, moon phases. No API key.
• io.github.davidmosiah/delx-agent-utilities — last commit 2026-08-06
  Stateless URL, DNS, x402, JWT, encoding, and parsing tools for AI agents.
• io.github.davidmosiah/delx-living-body — last commit 2026-08-06
  Meta-MCP that auto-detects installed wellness connectors and composes them into one body data layer.
• Delx MCP Server — last commit 2026-08-04
  MCP stdio bridge for the Delx Protocol: continuity, witness, recovery, ontology & passport tools
• io.github.davidmosiah/delx-memory — last commit 2026-08-06
  Local-first persistent memory MCP: shared SQLite key/value store, searchable, TTL-aware, secret-safe
• io.github.davidmosiah/eight-sleep-mcp — last commit 2026-08-05
  Unofficial Eight Sleep MCP server: read sleep trends, temperature, alarms; gated pod control.
• io.github.davidmosiah/fitbitmcp — last commit 2026-08-06
  Privacy-first, unofficial Fitbit MCP server for AI health, sleep, activity and heart-rate agents.
• io.github.davidmosiah/garminmcp — last commit 2026-08-06
  Privacy-first unofficial Garmin Connect MCP for AI health, sleep, activity & recovery agents.