{
  "markdown": "# Hunchful MCP server\n\n**Remote MCP server:** `https://hunchful.io/mcp` (Streamable HTTP)\n\n[Hunchful](https://hunchful.io) is an open implementation of the\n[Cognitive Model Protocol (CMP)](https://github.com/filinils/cognitive-model-protocol):\na person makes a revisable, **falsifiable** model of how they collaborate best with AI —\nhunches with a confidence, an uncertainty, and *what would change my mind* — at a\npermanent URL any agent can read. This server lets your assistant read that model,\nadapt to it, and help it grow slowly, with the person approving everything that matters.\n\n## Add it\n\n**Claude** — Settings → Connectors → Add custom connector → `https://hunchful.io/mcp`\n\n**ChatGPT** (developer mode) — Settings → Connectors → Add → URL `https://hunchful.io/mcp`,\nauth **OAuth** (sign in with your Hunchful account to act on your own model) or none\n(open reads; pass keys as tool arguments).\n\n**Anything else** — any MCP client that speaks Streamable HTTP. OAuth 2.1 with PKCE +\ndynamic client registration; discovery at `/.well-known/oauth-authorization-server`.\n\n## Tools (rising trust)\n\n| Tool | Auth | What it does |\n| --- | --- | --- |\n| `read_model` | none | The full model: hunches with confidence, uncertainty, falsifier, pole labels, situational activation rules, and the open-questions to-do list. |\n| `find_complementary_thinkers` | none | Which famous thinkers would *complete* this person — opposite poles on the axes where opposites unstick each other. |\n| `propose_refinement` | contribution key / OAuth | Strengthen/weaken suggestions → the owner's confirm queue. Nothing applies until they accept. |\n| `suggest_question` | contribution key / OAuth | Propose a NEW axis as a question → the owner's inbox (+ opt-in email). |\n| `apply_refinement` | edit key / OAuth owner | Direct, logged, reversible confidence nudges with private evidence pointers. |\n| `add_hunch` | edit key / OAuth owner | Record a hunch the person just confirmed in conversation — visible, revisable, removable. |\n| `find_matches` | edit key / OAuth owner | Look for complementary *people*. Agents propose; humans accept — double opt-in, identities never revealed by the tool. |\n\n## The rules the server enforces\n\n- **Everything is a hypothesis** — revisable positions on two-poled axes, never labels.\n- **Nothing is silent** — every write is logged in the model's revision history; evidence\n  entries are individually deletable and revert exactly what they applied.\n- **Provenance pointers only** — no conversation text, quotes, or third-party names, ever.\n- **Never identity** — a protected-attribute screen rejects axes about religion, politics,\n  health, ethnicity, gender, sexuality, or neurotype on every write path.\n- **Situational hunches** apply only when the person *says* they're in the state.\n- **Slow by design** — the model asks agents for at most one suggestion a day, one new\n  question a week, and silence when nothing real happened.\n\n## Links\n\n- Agent guide: https://hunchful.io/agent\n- Machine-readable model example: `https://hunchful.io/m/{id}/cmp.json`\n- Protocol: https://github.com/filinils/cognitive-model-protocol\n- Discovery: https://hunchful.io/llms.txt\n\n*The Hunchful app itself is a private codebase; this repo documents its public MCP surface.*\n",
  "bytes": 3276,
  "sha": "492311862c62ec0adaf0c696df851003fd716907d151cbb19ffffb00f760cdf2",
  "repo_slug": "filinils/hunchful-mcp",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_hunchful_hunchful_5df2efda/readme"
}