Logseq MCP Server

Logseq MCP Tools – Bridging AI Assistants and Your Local Knowledge Graph

The Logseq MCP server turns a locally‑hosted Logseq knowledge base into an AI‑friendly data source. By exposing Logseq’s REST API through the Model Context Protocol, it lets Claude or other MCP‑capable assistants read, create, update, and delete pages and blocks directly from within a conversation. This solves the long‑standing friction of having to switch between a note‑taking app and an AI tool, enabling seamless, real‑time interaction with personal knowledge.

At its core, the server offers a rich set of page and block functions under the namespace. Developers can query every page () or fetch a specific one by name, then drill down into its blocks with or target individual blocks via ID. Creation and manipulation are equally straightforward: , , , , and even bulk operations like or . The API also supports searching blocks by query, which is invaluable for quickly surfacing relevant notes during a dialogue.

The toolset is tailored to Logseq’s conventions, such as journal pages. By simply naming a page with the “mmm dth, yyyy” format (e.g., Apr 4th, 2025), the server automatically flags it as a journal page and assigns the correct attributes (, ). This eliminates boilerplate and lets AI assistants generate or retrieve daily logs without manual metadata handling. Additionally, block operations respect Logseq’s hierarchical structure—each block carries parent, level, and left relationships—so the assistant can maintain document organization while editing.

In practice, this MCP server powers workflows where an AI acts as a personal research assistant or note‑taking aide. A user can ask Claude to “create a new project page and add an introductory block,” or “search my notes for all mentions of ‘machine learning’ and summarize the findings.” The assistant can then instantly push results back into Logseq, keeping all information in a single, searchable graph. For developers building custom agents, the server’s clear namespace and parameterization make it easy to compose complex sequences of actions—e.g., generate a mind map, populate it with linked pages, and then export the structure.

Unique advantages stem from the tight integration with Logseq’s native features: automatic bullet rendering, double‑bracket page links, and hierarchical block data. Because the server operates locally, privacy is preserved—no sensitive content leaves the user’s machine. Moreover, developers can extend the MCP configuration to include environment variables for API URL and token, ensuring secure, repeatable deployments across tools like Cursor or Claude Code. Overall, Logseq MCP Tools transforms a local knowledge graph into an AI‑ready workspace, enabling developers to build smarter, contextually aware assistants that can read, write, and organize notes on demand.