This project provides an MCP server enabling AI assistants like Claude to interact directly with LogSeq graphs. It facilitates AI-driven knowledge management, allowing users to read, write, and manage notes seamlessly, transforming their knowledge base into an AI-powered workspace without context switching.

How It Works

The server acts as an intermediary, exposing LogSeq's Local HTTP API to AI clients via the uv package manager. It provides 16 core tools for comprehensive interaction, including page/block manipulation and search. A key feature is optional semantic vector search using local Ollama embeddings and LanceDB, ensuring data privacy. Version 1.1.0+ introduces smart markdown parsing, automatically converting markdown frontmatter, headings, and lists into LogSeq's native block structure. It also supports Logseq's DB-mode graphs for property management.

Quick Start & Requirements

Integration involves adding the server configuration to Claude Desktop or using the claude mcp add mcp-logseq command. Prerequisites include a running LogSeq instance with its HTTP APIs server enabled and an API token generated. The uv Python package manager is required. Optional setup for vector search involves installing Ollama. Links to detailed guides for vector search (VECTOR_SEARCH.md) and development (DEVELOPMENT.md) are mentioned.

Highlighted Details

• Extensive Toolset: Offers 16 core tools for LogSeq content management, plus 3 optional vector search tools.
• Smart Markdown Parsing (v1.1.0+): Automatically translates markdown frontmatter, headings, lists, code blocks, and checkboxes into LogSeq's structured block format.
• Privacy-Preserving Vector Search: Enables semantic, meaning-based search using local AI embeddings (Ollama) and LanceDB, with no data leaving the user's machine.
• DB-Mode Support: Allows interaction with Logseq DB-mode graphs, including reading and writing class properties on blocks.

Maintenance & Community

The provided README does not contain specific details regarding notable contributors, sponsorships, or community channels.

Licensing & Compatibility

The README does not specify a software license, presenting ambiguity regarding usage rights and commercial compatibility.

Limitations & Caveats

DB-mode graph support is noted as opt-in and beta. Vector search requires additional dependencies (Ollama, LanceDB). The absence of a stated license is a critical caveat. Some features are specific to Logseq DB-mode graphs.