Claix MCP

Claix as an MCP server

Endpoint

MCP https://claix.dev/mcp

Claix exposes a Model Context Protocol (MCP) server (version 1.13.0, 32 tools) so AI assistants and automation tools can run Retrieve first (routes, agents, live database ask), plus extraction, Agent mode, Markdown, and multi-document Q&A without hand-written REST integrations.

The public server URL is https://claix.dev/mcp. Auth uses your Claix API key in the x-api-key header (or the optional api_key tool argument if the client cannot send custom headers).

Tools use dot notation (claix.schemas.list, claix.extract.pdf, …) and return structuredContent with success, data, and error.

1. Compatible clients

Any MCP-capable program can connect to Claix. Common production and development clients:

• Cursor
• Claude Desktop
• Claude Code
• Windsurf
• Cline
• Continue
• Zed
• Smithery
• Lovable
• Replit Agent
• ChatGPT (with MCP connectors)
• n8n
• Make
• Zapier (via HTTP/MCP)
• LangGraph / LangChain
• Any MCP client with Streamable HTTP or SSE

IDEs and coding agents (Cursor, Windsurf, Cline…) usually take JSON config with the server URL and headers. Smithery and Streamable HTTP clients send JSON-RPC directly to POST /mcp. Claude Desktop often uses SSE via mcp-remote (GET /mcp + POST /mcp/message).

2. Authentication

Recommended — HTTP header on every request:

x-api-key: <YOUR_API_KEY>

In Smithery, configure the header with x-from: { "header": "x-api-key" } so the user pastes their key when connecting. The API key is optional in the config schema if the client already sends x-api-key.

Per-tool fallback: many tools accept api_key in arguments if the MCP client cannot send custom headers.

3. Transport modes

Mode	Typical use	How to connect
Streamable HTTP	Smithery, Cursor, modern clients	POST /mcp with JSON-RPC (initialize, tools/list, tools/call). Header Accept: application/json, text/event-stream
SSE legacy	Claude Desktop, mcp-remote	GET /mcp (SSE stream) + POST /mcp/message?sessionId=… for messages

Both modes are active on the same base URL. The server detects the request type automatically.

Static metadata for scanners (Smithery, clients): https://claix.dev/.well-known/mcp/server-card.json

5. Markdown, multi-document, and Retrieve

Area	Tools	Notes
Markdown	claix.markdown.*	No schema_id; optional save → row_id
Multi-document	claix.multi_document.ask	Multipart field "file" (up to 10) + questions JSON
Retrieve	claix.retrieve.ask, claix.routes.*, claix.agents.*	POST /retrieve/{agent_name}

6. Available tools

After connecting, call tools/list for all 32 tools with inputSchema, outputSchema, and annotations:

Tool	Description
claix.schemas.list	Lists account schemas (id, name, type, is_agent_mode, agent_definition). Call first to discover schema_id.
claix.schemas.get	Gets one schema by schema_id (GET /api/get-schema/{schema_id}) with its full definition, including Guardar Datos (window_context). Free call.
claix.schemas.create	Creates a schema (POST /api/create-schema): name, type, schema_definition; optional Agent mode, window_context (Save Data → insert mapped row; default false), and source verification.
claix.schemas.update	Updates a schema (PUT /api/update-schema/{schema_id}): same body as create; window_context = Save Data. Free call.
claix.schemas.delete	Deletes an account schema (POST /api/delete-schema) by schema_id.
claix.extract.excel	Excel/CSV → typed JSON (POST /api/excel-json). First sheet. file + schema_id (no space_id). If Guardar Datos (window_context), may return row_id. Returns log_id.
claix.extract.pdf	PDF → typed JSON (POST /api/pdf-json). Text or scanned, max 15 MB. file + schema_id (no space_id). If Guardar Datos, may return row_id. Returns log_id.
claix.extract.doc	Document (.docx,.txt,.md,.rtf) → JSON (POST /api/doc-json). file + schema_id (no space_id). If Guardar Datos, may return row_id. Returns log_id.
claix.extract.image	Image (JPEG, PNG, WebP, HEIC) → JSON (POST /api/img-json). file + schema_id (no space_id). If Guardar Datos, may return row_id. Returns log_id.
claix.extract.text	Text/HTML/XML → JSON (POST /api/txt-json). content field, no file. schema_id (no space_id). If Guardar Datos, may return row_id. Returns log_id.
claix.extract.audio	Audio (MP3, WAV, M4A, OGG) → JSON (POST /api/audio-json). file + schema_id (no space_id). If Guardar Datos, may return row_id. Returns log_id.
claix.convert.json_to_excel	JSON → Excel.xlsx (POST /api/json-excel). Returns file_base64.
claix.agent.excel	Excel/CSV + Agent mode → data[], agent_data, and log_id (POST /agent/excel-json). file + schema_id (no space_id). If Guardar Datos, may return row_id.
claix.agent.pdf	PDF + Agent mode → data[], agent_data, and log_id (POST /agent/pdf-json). file + schema_id (no space_id). If Guardar Datos, may return row_id.
claix.agent.doc	Document + Agent mode (POST /agent/doc-json). data[], agent_data, and log_id. file + schema_id (no space_id). If Guardar Datos, may return row_id.
claix.agent.image	Image + Agent mode (POST /agent/img-json). data[], agent_data, and log_id. file + schema_id (no space_id). If Guardar Datos, may return row_id.
claix.agent.text	Text/HTML/XML + Agent mode (POST /agent/txt-json). content + schema_id (no space_id). If Guardar Datos, may return row_id.
claix.agent.audio	Audio + Agent mode (POST /agent/audio-json). data[], agent_data, and log_id. file + schema_id (no space_id). If Guardar Datos, may return row_id.
claix.markdown.excel	Excel/CSV → Markdown (POST /api/excel-markdown). No schema_id. save=true inserts with Markdown rules and may return row_id.
claix.markdown.pdf	PDF → Markdown (POST /api/pdf-markdown). No schema_id. Optional save → row_id.
claix.markdown.doc	Document → Markdown (POST /api/doc-markdown). No schema_id. Optional save → row_id.
claix.markdown.image	Image → Markdown (POST /api/img-markdown). No schema_id. Optional save → row_id.
claix.markdown.text	Text/HTML/XML → Markdown (POST /api/txt-markdown). content field. Optional save → row_id.
claix.markdown.audio	Audio → Markdown (POST /api/audio-markdown). No schema_id. Optional save → row_id.
claix.multi_document.ask	Typed questions across up to 10 files (POST /multi-document). Multipart: repeatable file field + questions JSON. Optional window_context/window_time/space_id (OpenAPI).
claix.retrieve.ask	Ask a Retrieve agent (POST /retrieve/{agent_name}). Body { questions[{ question, format? }] }; format required only for razonada mode.
claix.routes.create	Create a Retrieve route (POST /routes): name, description, rules, tables[{ name, columns }].
claix.routes.edit	Edit a Retrieve route (PUT /routes/edit/{route_id}). Same body as create + route_id.
claix.routes.delete	Delete a Retrieve route (POST /routes/delete/{route_id}).
claix.agents.create	Create a Retrieve agent (POST /agents): name, slug, description, prompt, routes[], response_mode bruto|razonada.
claix.agents.edit	Edit a Retrieve agent (PUT /agents/edit/{agent_id}). Same body as create + agent_id.
claix.agents.pause	Pause or activate an agent (POST /agents/pause/{agent_id}). Body { status: paused|active }.

7. Workflow prompts

The server exposes reusable prompts via prompts/list:

Prompt	Description
workflow.discover-and-extract	Discover schemas via claix.schemas.list and extract JSON with claix.extract.*.
workflow.agent-document-analysis	Run claix.agent.* when the schema has is_agent_mode enabled.
workflow.invoice-pdf	Invoice PDF workflow using claix.extract.pdf.
workflow.markdown-save	Convert to Markdown with claix.markdown.* and optionally persist with save=true (row_id).
workflow.multi-document-ask	Ask typed questions across several files with claix.multi_document.ask (no knowledge spaces).
workflow.retrieve-ask	Query a Retrieve agent with claix.retrieve.ask after configuring routes and agents.

8. Resources

Documentation resources available with resources/list:

URI	Description
claix://docs/mcp	MCP connection guide and tool catalog.
claix://docs/openapi	Claix OpenAPI specification URL.
claix://docs/tools	JSON catalog of all MCP tools.

9. Smithery example

One-click install:

npx -y @smithery/cli run info-f4xz/claix

Manual config:

{
  "mcpUrl": "https://claix.dev/mcp",
  "headers": {
    "x-api-key": "YOUR_API_KEY"
  }
}

Directory: smithery.ai/server/info-f4xz/claix

10. Cursor example

Add to Cursor MCP settings (Settings → MCP):

{
  "mcpServers": {
    "claix": {
      "url": "https://claix.dev/mcp",
      "headers": {
        "x-api-key": "YOUR_API_KEY"
      }
    }
  }
}

11. Claude Desktop example

{
  "mcpServers": {
    "claix": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://claix.dev/mcp",
        "--header",
        "x-api-key:YOUR_API_KEY"
      ]
    }
  }
}

12. n8n and automation

In n8n use an HTTP Request node with POST https://claix.dev/mcp, JSON-RPC headers, and body {"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"claix.extract.pdf","arguments":{"schema_id":"…","file_base64":"…"}}}. Also try claix.markdown.pdf, claix.multi_document.ask, or claix.retrieve.ask. Community MCP nodes work too: point the base URL at Claix and pass x-api-key.