MCP Directory

How to add Graphlit MCP Server to Claude Desktop

Ingest, search, and RAG over data from Slack, Discord, Google Drive, GitHub, Jira, Linear and more via Graphlit. Paste the config into ~/Library/Application Support/Claude/claude_desktop_config.json and restart Claude Desktop.

Last updated June 14, 2026 · 375★ · stdio · apikey

Claude Desktop config for Graphlit MCP Server

npx -y graphlit-mcp-server
{
  "mcpServers": {
    "graphlit-mcp-server": {
      "command": "npx",
      "args": [
        "-y",
        "graphlit-mcp-server"
      ],
      "env": {
        "GRAPHLIT_ORGANIZATION_ID": "your-organization-id",
        "GRAPHLIT_ENVIRONMENT_ID": "your-environment-id",
        "GRAPHLIT_JWT_SECRET": "your-jwt-secret"
      }
    }
  }
}

Setup steps

  1. 1Open Claude Desktop → Settings → Developer → Edit Config (this opens ~/Library/Application Support/Claude/claude_desktop_config.json).
  2. 2Paste the Graphlit MCP Server config below under the top-level "mcpServers" key.
  3. 3Fill in any placeholder secrets (API keys, paths) in the snippet.
  4. 4Save the file, then fully quit and reopen Claude Desktop.
  5. 5Open a chat and confirm Graphlit MCP Server's tools appear under the 🔌 tools menu.

Before you start

  • Node.js 18.x or higher
  • An active Graphlit Platform account (portal.graphlit.dev) with access to the API settings dashboard
  • GRAPHLIT_ORGANIZATION_ID, GRAPHLIT_ENVIRONMENT_ID and GRAPHLIT_JWT_SECRET from the API settings dashboard

What Graphlit MCP Server can do in Claude Desktop

Query Contents

Search and query ingested content in the Graphlit project.

Query Collections

Query collections of content.

Query Feeds

Query ingestion feeds.

Query Conversations

Query stored conversations.

Retrieve Relevant Sources

Retrieve the most relevant content sources for a query.

Retrieve Similar Images

Retrieve images similar to a given image.

Visually Describe Image

Generate a visual description of an image.

Prompt LLM Conversation

Run a RAG conversation with an LLM grounded in ingested content.

Security

Requires Graphlit credentials (GRAPHLIT_ORGANIZATION_ID, GRAPHLIT_ENVIRONMENT_ID, GRAPHLIT_JWT_SECRET) set as environment variables. Optional per-connector credentials (Slack bot token, Discord bot token, GitHub PAT, Notion API key, Google OAuth tokens, Linear/Jira keys, etc.) can also be supplied as env vars; store all secrets securely and avoid committing them.

Graphlit MCP Server + Claude Desktop FAQ

Where is the Claude Desktop config file?

Claude Desktop reads MCP servers from ~/Library/Application Support/Claude/claude_desktop_config.json. Paste the Graphlit MCP Server config there under the "mcpServers" key and restart the client.

Is Graphlit MCP Server safe to use with Claude Desktop?

Requires Graphlit credentials (GRAPHLIT_ORGANIZATION_ID, GRAPHLIT_ENVIRONMENT_ID, GRAPHLIT_JWT_SECRET) set as environment variables. Optional per-connector credentials (Slack bot token, Discord bot token, GitHub PAT, Notion API key, Google OAuth tokens, Linear/Jira keys, etc.) can also be supplied as env vars; store all secrets securely and avoid committing them.

Which credentials are required?

Only GRAPHLIT_ORGANIZATION_ID, GRAPHLIT_ENVIRONMENT_ID and GRAPHLIT_JWT_SECRET are required. Per-connector credentials (e.g. Slack, Discord, GitHub, Notion, Google, Linear, Jira tokens) are optional and only needed for those data connectors.

Do I need separate web scraping or search tools?

No. Web crawling and web search (including podcast search) are built in as MCP tools, so you don't need to integrate Firecrawl, Exa or similar separately.

Which MCP clients are supported?

It works with any MCP client, with documented setup for Cursor, Windsurf, Cline, VS Code, and Claude Desktop (via Smithery). The README also highlights Goose.

View repo Full Graphlit MCP Server page