{
  "markdown": "# clutter-mcp — synthetic data & synthetic document generator (MCP server)\n\nA [Model Context Protocol](https://modelcontextprotocol.io) **stdio** server that lets an AI agent\ndrive the whole [Clutter](https://clutter.run) pipeline — invent a believable synthetic company,\nmass-produce the documents, spreadsheets, emails, images and datasets it would really have, poll for\ncompletion, and fetch download URLs — through Clutter's public REST API, authenticated with an API key.\n\nUse it to fill dev/test/demo systems (SharePoint, CRMs, file shares) with realistic content —\n**synthetic documents and synthetic data** that stand in for the real thing. Generate **test data,\ndemo data and training data**, or **test documents and training documents**, on demand — or give an\nAI agent believable data to reason over, all without touching real or sensitive data.\n\nIt's a thin HTTPS client with **no dependency on the rest of the Clutter codebase**.\n\n## Quick start\n\n1. Create a free account at **https://clutter.run**, open **Settings**, and mint an API key\n   (`clt_live_…`, shown once).\n2. Add the server to your MCP client config (Claude Code, Claude Desktop, etc.):\n\n```json\n{\n  \"mcpServers\": {\n    \"clutter\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"clutter-mcp\"],\n      \"env\": {\n        \"CLUTTER_API_KEY\": \"clt_live_…\"\n      }\n    }\n  }\n}\n```\n\nThat's it — `CLUTTER_API_URL` defaults to `https://clutter.run/api`, so only the key is required.\n\n## Configuration\n\n| Env var           | Required | Default                   | Notes                                          |\n| ----------------- | -------- | ------------------------- | ---------------------------------------------- |\n| `CLUTTER_API_KEY` | yes      | —                         | `clt_live_…` key (web app → **Settings**)      |\n| `CLUTTER_API_URL` | no       | `https://clutter.run/api` | Override only to target another deployment      |\n\n## Run directly\n\n```bash\nCLUTTER_API_KEY=clt_live_… npx -y clutter-mcp\n```\n\n## Typical agent flow\n\n1. `build_org` — describe a company in a sentence → returns an `orgId` (omit `projectId` to\n   auto-create a project). Free.\n2. `wait_for_org` — block until the company build is `ready`.\n3. `create_run` — generate content against the company:\n   - `doc_generator` → a batch of documents (`docx`/`pdf`/`xlsx`/`eml`/`jpg`), folder-organised.\n   - `data_generator` → one tabular dataset (`xlsx`/`csv`/`json`) with an exact row count.\n   - `doc_metadata_gen` → one metadata record per document of a prior doc run.\n4. `wait_for_run` — block until the run is `complete`.\n5. `list_run_documents` + `get_document_url`, or `build_zip` + `get_zip_url` for the whole run as a\n   single ZIP (folder tree preserved — ready to drop into SharePoint or a file share).\n\n## Tools\n\n- **Identity / usage:** `clutter_whoami`, `get_usage`\n- **Projects:** `list_projects`, `create_project`, `delete_project`\n- **Companies:** `list_orgs`, `build_org`, `get_org`, `wait_for_org`, `query_org`, `delete_org`\n- **Runs:** `list_runs`, `create_run`, `get_run`, `wait_for_run`, `list_run_documents`,\n  `get_document_url`, `build_zip`, `get_zip_url`, `delete_run`\n\n`build_org` and `create_run` are asynchronous (return an id immediately); use the `wait_for_*` tools\nto block until a terminal state, or poll `get_org` / `get_run`. Each tool maps to a REST endpoint and\nreturns the raw JSON response.\n\n## Billing\n\nCompany builds, metadata and \"ask the company\" are free. New accounts get 10 free documents + 100\nfree data rows, then pay-as-you-go (all prices USD: $0.10/document, $0.06/10 data rows). A `402` from `create_run`\nmeans insufficient credit — top up at https://clutter.run/billing.\n\n## Loading content into SharePoint\n\nClutter hands you download URLs; your agent does the upload (via Microsoft Graph). Beyond a\nplain \"drop files into a library\", the generated metadata can drive automation — apply\nsensitivity labels, set permissions, assign retention, route flat files to the right\nsite/library by metadata, or even generate the information architecture itself. Worked\npatterns with Graph calls: **https://clutter.run/sharepoint-cookbook.md**\n\n## Reference\n\nFull REST API: **https://clutter.run/api/docs** (Swagger UI) · machine-readable guide:\n**https://clutter.run/llms.txt**\n\n## License\n\nMIT\n",
  "bytes": 4283,
  "sha": "a783d0b860253dde55a81de53390a621ad2ddbc93a47c18dc67fa2b5cc09593c",
  "repo_slug": "michaelhologram/clutter-mcp",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_github_michaelhologram_clutter_mcp_c9c3f793/readme"
}