Graphlit MCP Server

Example Use Cases

• 1

  Ingest and centralize content from Slack, Discord, Google Drive, email, Jira, GitHub, and other tools into a searchable Graphlit project.

• 2

  Perform retrieval-augmented generation (RAG) and retrieve relevant sources, collections, feeds or conversations during LLM interactions.

• 3

  Built-in web crawling/search, multimedia transcription, and publishing (audio/image) pipelines without wiring separate toolchains.

Description

The Graphlit MCP Server lets MCP clients (like Cursor, Windsurf, Goose, and Cline) integrate with the Graphlit service to ingest documents, messages, emails, and more into a searchable project. It extracts and transcribes content (PDFs, DOCX, PPTX, HTML, audio/video) to Markdown or text, provides built-in web crawling and search tools, and exposes retrieval, RAG, extraction, publishing and operations endpoints. It supports many data connectors (Slack, Discord, Google Drive, GitHub, Jira, Notion, etc.) to centralize knowledge for RAG workflows and agent use. The server is configurable via environment variables and can be run via NPX or integrated into MCP clients' server configs.