mcp server
Incisory
Audit whether ChatGPT, Gemini & Perplexity recommend a business — verbatim proof, then fix it.
Description as published by the maintainer. Source
- version 1.0.1
- active
active — Registry entry last updated 2026-07-10.
What this server can do
23 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.
ai_traffic(days, token)- Show how many visits AI assistants sent the client (last 30 days) + value, and the one-line tracking snippet to install.
audit_visibility(niche, token, website, business)- Run an AI-visibility audit for a business (requires an Incisory account — every account includes one free audit, which runs at signup; paid accounts can re-run freely). Pass just the website and Incisory reads the site to work out what it sells. Asks ChatGPT, Gemini and Perplexity the questions its buyers ask; returns visibility %, verbatim answers (word for word), who AI recommends instead, and the 3 fastest fixes. Takes ~60s. Required: website.
brand_accuracy(token)- The latest brand-accuracy scan: what AI assistants currently state about the business, with each claim checked against the real web (correct / outdated / likely-wrong / missing) and the highest-leverage fix. Runs weekly on the paid plan.
changelog(days, token)- List the automated work Incisory has done for the account (scans, articles written, Reddit threads found) over the last N days.
citation_gap(niche, token, website, business)- Show which third-party sources (review sites, directories, communities, Wikipedia, etc.) ChatGPT/Gemini actually cited while answering questions where you were NOT named — a prioritized list of exactly where to get listed, reviewed, or mentioned next. Use your account token (uses your latest scan), or pass business/website/niche for an ad-hoc audit.
competitor_decode(token)- The latest competitor decode for the account: why AI engines cite the top rival instead of you — their footprint (matchable vs out of reach) and the specific moves you can make this quarter. Runs weekly on the paid plan.
find_reddit_opportunities(niche, token, website, business)- Search for real, high-intent Reddit threads relevant to a business and draft a disclosed, helpful reply for each — queued for human review, never auto-posted (posting is your call, from your own account). Use your account token, or pass business/website/niche directly for an ad-hoc search.
get_audit(id)- Fetch a previously run audit by its id (from the shareable result URL). Required: id.
google_data(token)- Real (not estimated) Google Search Console clicks/impressions/ranking-position and GA4 sessions, plus verified AI-referral session counts, for your connected Google account. If not connected yet, returns the link to connect. Uses your account token.
list_content_queue(token)- List generated article drafts (title, target question, status) for your account.
list_earned_mentions(token)- List pending outreach pitches (white-hat, drafted for you to send) to the sources AI already cites instead of your account — never auto-sent.
list_prompts(token)- The questions Incisory re-asks the AI engines every week for your account, with each engine's latest verdict.
list_reddit_queue(token)- List pending Reddit opportunities (copy-paste drafts + thread links) for your account.
mark_content(id, token, status, published_url)- After you publish a draft yourself (or decide to skip it), mark it so it leaves the queue. Required: id, status.
mark_earned_mention(id, token, status)- After you send a drafted pitch yourself (or decide to skip it), mark it here so it stops appearing in your queue. Required: id, status.
mark_reddit_opportunity(id, token, status)- After you post a drafted reply yourself (or decide to skip it), mark it here so it stops appearing in your queue. Required: id, status.
my_report(token)- Full latest weekly report for a subscribed client (verbatim answers, quick wins) using your account token.
my_visibility(token)- For a subscribed Incisory client: latest AI-visibility snapshot + trend, using your account token.
set_autonomy(token, autonomy)- Choose how much Incisory does on its own each week: "notify" (just tell me), "approve" (draft and wait for me), or "auto" (do it all). Required: autonomy.
track_prompt(text, token)- Add a question your buyers ask (e.g. "best CRM for small law firms") to your tracked set — it gets re-asked across ChatGPT/Gemini/Perplexity every weekly scan so you see a real per-question trend. Max 12 tracked questions. Required: text.
upgrade_plan- Get the Stripe checkout link to subscribe — one plan, €49/mo, everything included, with a free 3-day trial.
visibility_history(token)- Visibility trend across your weekly snapshots. Uses your account (OAuth or token) — you can only read your own history.
write_article(niche, token, website, business, target_prompt)- Generate a genuinely good, branded article designed to get cited by AI assistants for a specific buyer question. Returns a draft (title + body + FAQ schema) stored in your content queue for review — not auto-published. Use your account token, or pass business/website/niche for an ad-hoc draft. Required: target_prompt.
Last successful function declaration observed on . Source: https://incisory.com/mcp. We list what the server declared; we do not call any of these functions.
Endpoint status observed on . Source: https://incisory.com/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 | 1.0.1 | Latest version string the maintainer published to the registry. | as of fetch | Model Context Protocol | |
| Registry record last updated | 2026-07-10 | When the registry record was last updated by its maintainer. | point in time | Model Context Protocol | |
| First listed in the MCP Registry | 2026-07-10 | 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 | 23 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 | incisory.com | |
| mcp endpoint status | ok | The server listed 23 functions when asked. | as of probe | incisory.com |
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