ZBS Index What actually exists in applied AI, with the source next to it

mcp server

litmus

Grade MCP servers A to F with the open behavioral litmus. npm: full toolset; hosted: lookups only.

Description as published by the maintainer. Source

  • version 0.36.0
  • active

active — Registry entry last updated 2026-07-20.

What this server can do

3 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.

check_server(server_ref)
Read a server's published behavioral grade (A-F) from polygraph.so in under a second; no execution. The pre-flight check before recommending or installing an MCP server. On a miss it returns not_available (unevaluated: neither safe nor unsafe) with next steps. Required: server_ref.
list_servers(grade, limit, offset)
Servers with a published polygraph grade, sorted A-first then by ref. Returns up to `limit` rows starting at `offset` (default limit 25, capped at 100 per call); `grade` restricts to one letter. `summary` always covers the full graded corpus (a total plus a count per grade), regardless of `grade`, `limit`, or `offset`.
request_grade(server_ref)
Record a grade request with polygraph.so. Recording is free; only real, plausibly-MCP targets are accepted. Grading starts once the request's one-time $1 USD fee is paid: payment.payUrl is the human/browser checkout (paid in $POLYGRAPH), and payment.x402Url is the agent rail, POST the same request body there with an x402-capable client. A bare POST to x402Url returns a 402 with the exact payment requirements; retry with an X-PAYMENT header. Asset is USDC on Base mainnet (eip155:8453). On the x402 rail settlement is deferred: the authorization is charged only once a grade lands, and a run the harness cannot complete voids it, so nothing is charged. Paying starts a 48h grading clock. The fee buys the run, never the grade. After paying, poll check_server with the same server_ref for the published result, or poll statusUrl for progress. Required: server_ref.

Last successful function declaration observed on . Source: https://polygraph.so/api/mcp. We list what the server declared; we do not call any of these functions.

Endpoint status observed on . Source: https://polygraph.so/api/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.

Signal Value What it measures Window Observed Source
Latest published version 0.36.0 Latest version string the maintainer published to the registry. as of fetch Model Context Protocol
Registry record last updated 2026-07-20 When the registry record was last updated by its maintainer. point in time Model Context Protocol
First listed in the MCP Registry 2026-07-20 Date this server was first published to the official MCP Registry. Not a usage or quality measure. point in time Model Context Protocol
mcp tools declared 3 tools Number of functions the server itself declared when asked to list them. This is what the server offers an agent, not a measure of how well any of them work. as of probe polygraph.so
mcp endpoint status ok The server listed 3 functions when asked. as of probe polygraph.so

Where to get it

This record as data

Every field on this page, with its source and observation date, is in the catalog JSON. Fetch the whole kind at once instead of parsing this HTML.

GET /api/v1/entries/mcp_server.json

Sources

  1. Tools declared by the MCP server at https://polygraph.so/api/mcp — polygraph.so, observed , trust tier 1.
  2. Official MCP Registry — Model Context Protocol, observed , trust tier 1.