Claude MCP Setup

Connect Claude-compatible MCP clients to ReadMe MCP for docs and the central TextMine MCP endpoint for product tools.

Use this setup when Claude or a Claude-compatible MCP client should work with TextMine docs, Public API v3 schemas, or TextMine MCP tools.

Which endpoint to use

NeedEndpoint
Read developer docs, schemas, and generated API-reference toolshttps://developers.textmine.com/mcp
Operate TextMine through central product MCP toolshttps://mcp.textmine.com/mcp

Use https://mcp.textmine.com/mcp for public TextMine MCP setup. Use environment-specific MCP hosts only for private preview, staging validation, or dedicated deployments when TextMine explicitly issues that host.

Remote connector setup

For Claude surfaces that support remote MCP connectors, configure the central TextMine MCP endpoint with the caller's TextMine API key:

Server URL: https://mcp.textmine.com/mcp
Authorization: Bearer tm_your_api_key

If the client owns or rewrites Authorization, use this header instead:

X-TextMine-Api-Key: tm_your_api_key

Add https://developers.textmine.com/mcp as a separate MCP connector when Claude should search developer docs or inspect schemas. Use the caller's own TextMine key for product-tool execution so Vaults, provider connections, users, and permissions resolve correctly.

If an autonomous third-party agent does not yet have a key, use Agent self-service signup. Signup is REST-first and returns an encrypted service-account key for the trusted runtime to store.

Claude Desktop note

Some Claude Desktop deployments still expect local stdio MCP servers in claude_desktop_config.json, while newer Claude surfaces can use remote connectors. If your Desktop install only accepts local commands, use a remote-MCP bridge approved by your organisation and point that bridge at https://mcp.textmine.com/mcp.

Keep the TextMine bearer token in an environment variable or the client's secure credential store rather than writing it into a shared JSON file.

First prompt to try

Use TextMine MCP to read the For Agents page. Then explain the safe operation sequence for generating a report from Vault documents, including which operations are discovery calls and which operation creates the report. Do not call write operations yet.

What Claude should do first

  • Read For Agents.
  • Use listVaults before Vault-scoped actions.
  • Use listIntegrationProviders, folder listing, and file listing before provider import.
  • Ask for confirmation before enabled workflows, scheduled agents, destructive actions, provider disconnects, or API-key creation.