{
  "markdown": "# Fractal Process Intelligence — Research Alpha\n\n**Fractal measures whether additional AI actually created additional intelligence.**\n\nFractal is not trying to make AI smarter. It observes research *process* and produces an X-Ray of what a polished final answer may hide.\n\n> Primary question: **Did additional AI actually create additional independent intelligence?**\n\nThis repository is the **public distribution surface** for the Fractal Research MCP (recorder/observer). The live remote endpoint is already hosted; this repo documents how to connect and give feedback.\n\n**Research Alpha:** This is an active research experiment, **not** a validated commercial product.\n\n---\n\n## What it does\n\nFractal Process Intelligence MCP is **recorder/observer only**.\n\nCompatible AI clients can record observable research-process artifacts:\n\n- participating agents  \n- meaningful sources  \n- claims  \n- uncertainty  \n- disagreement  \n- tool usage  \n\nThen `finish_run` compiles a process **X-Ray** (hosted report URL).\n\nPublic product page: [fractalinfinity.io/research-lab.html](https://fractalinfinity.io/research-lab.html)  \nSample X-Ray: [fractalinfinity.io/report/sample.html](https://fractalinfinity.io/report/sample.html)\n\n---\n\n## What it does NOT do\n\n- Call AI models (no Anthropic / OpenAI / Gemini inference)  \n- Buy model inference on your behalf  \n- Require provider API keys  \n- Inspect hidden chain-of-thought  \n- Automatically spawn agents  \n- Estimate Claude/OpenAI billing or token dollar cost  \n\nHealth check advertises `\"inference\":\"none\"`.\n\n---\n\n## Public MCP endpoint\n\n```text\nhttps://mcp.fractalinfinity.io/mcp\n```\n\nHealth: https://mcp.fractalinfinity.io/health\n\n### Current recorder tools\n\n| Tool | Purpose |\n|------|---------|\n| `fractal_start_run` | Open a run |\n| `fractal_record_agent` | Record agent identity / optional metadata |\n| `fractal_record_source` | Record a cited source + lineage |\n| `fractal_record_claim` | Record claim stance |\n| `fractal_record_tool` | Record tool-call evidence |\n| `fractal_finish_run` | Finalize → X-Ray URL |\n\n---\n\n## Claude Desktop (Windows) alpha setup\n\n1. Claude Desktop → **Settings → Developer → Edit Config**  \n2. Merge the MCP entry below (do **not** wipe other config keys).  \n3. Save.  \n4. Fully quit Claude (system tray → Quit).  \n5. Reopen Claude.  \n6. Settings → Developer should show `fractal-process-intelligence` running.\n\n```json\n{\n  \"mcpServers\": {\n    \"fractal-process-intelligence\": {\n      \"command\": \"C:\\\\Program Files\\\\nodejs\\\\npx.cmd\",\n      \"args\": [\n        \"-y\",\n        \"mcp-remote\",\n        \"https://mcp.fractalinfinity.io/mcp\"\n      ]\n    }\n  }\n}\n```\n\nSee also: [`examples/claude-desktop.mcp.fragment.json`](examples/claude-desktop.mcp.fragment.json)\n\nMicrosoft Store builds of Claude may store config under a package LocalCache path; prefer **Edit Config** so Claude opens the file it actually reads. Use the full `npx.cmd` path on Windows.\n\n---\n\n## Example research instruction\n\nPaste into a new research conversation after connecting:\n\n```text\nUse Fractal Process Intelligence to record this research run.\n\nRecord the participating agents, meaningful sources, important claims,\nuncertainty, disagreement and significant tool usage.\n\nFinish the run and return the Fractal X-Ray URL.\n\nDo not interpret the X-Ray for me before I open it.\n```\n\nThen ask your normal research question. Read the AI’s final answer first. Open the X-Ray second.\n\n---\n\n## Privacy / security boundary\n\nFractal records **only** artifacts explicitly sent through its MCP tools.\n\nIt does **not** automatically record every Claude (or other) conversation. For normal chats, do not start a Fractal run / do not invoke its tools.\n\n---\n\n## Feedback (the research question)\n\nOpen a [GitHub Issue](https://github.com/chevy155/fractal-process-intelligence-mcp/issues/new?template=research_alpha_feedback.md) or reply in discussions.\n\nPlease answer:\n\n1. **Did the X-Ray reveal something consequential you would not have known from the final AI answer alone?**  \n2. **Would you want this attached to your normal research or agent workflow?**  \n\nOptional: what was confusing or misleading?\n\n---\n\n## Links\n\n| Resource | URL |\n|----------|-----|\n| Research Lab | https://fractalinfinity.io/research-lab.html |\n| Sample X-Ray | https://fractalinfinity.io/report/sample.html |\n| MCP endpoint | https://mcp.fractalinfinity.io/mcp |\n| Health | https://mcp.fractalinfinity.io/health |\n\n---\n\n## License\n\nMIT — see [LICENSE](LICENSE).\n",
  "bytes": 4471,
  "sha": "d6d2d0a8b9574486dd44d0c9bbb78b8abcc110d9705a6e8f9a7bc3e59fb3c58e",
  "repo_slug": "chevy155/fractal-process-intelligence-mcp",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_github_chevy155_fractal_process_intel_d332c8ad/readme"
}