Graphlit

What it does

Graphlit turns your fragmented developer and product management data into a searchable, RAG-ready knowledge base. It ingests content from a vast array of sources—including Slack, Discord, GitHub, Jira, and Google Drive—and converts documents and web pages into Markdown for precise AI retrieval.

Tools

• Query Contents/Collections/Feeds: Search and retrieve specific knowledge from your ingested data.
• Retrieve Relevant Sources: Pull the most pertinent documents for a given query.
• Web Crawling & Search: Built-in web discovery and mapping without needing external scrapers.
• Ingestion Tools: Directly add files, web pages, and messages from supported connectors.
• Data Connectors: Specialized tools for Notion, Reddit, Linear, and various email providers.

Installation

Add the following to your claude_desktop_config.json:

{
  "mcpServers": {
    "graphlit": {
      "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"
      }
    }
  }
}

Supported hosts

Confirmed support for Claude Desktop, Windsurf, Cursor, and Goose.