Labsco
MCP SERVER

Logseq MCP Server

by dailydaniel ↗

Let agents read your whole Logseq graph but write only inside their own namespace, with every write logged to today's journal.

Agent Memory✓ Verified
Summary
Turns an existing Logseq graph into agent memory you can still supervise, because the write surface is small and every write leaves a trace.

Two design choices do the work: agents write only under their own prefix, and each successful write appends a line to today's journal, so reviewing what they did is part of a note you already open. The guide tool and the named-query mechanism are the other half — they hand a model verified Datalog instead of letting it compose queries by hand.

What it is

A server for the file/Markdown version of Logseq, reached through Logseq's local HTTP API. Reads span the graph; writes are confined to one namespace prefix.

What you get
  • Full-text search across block content, using ripgrep over the graph folder when you point it at one, otherwise a datascript content match
  • Task lookup by marker, tag, descendant tag, page or priority
  • Namespace listing, which surfaces the child pages a parent page's own read will never show
  • Pages and blocks read as a normalized tree with ((block refs)) resolved, keeping each resolved block's uuid and status so you can act on it
  • Raw Datalog when you need it, plus named queries from the config that can be registered as tools of their own
  • A built-in guide tool returning the verified query and write rules for the graph, so a model does not re-derive them
  • Writes inside the agent prefix — create or update a page, set or remove page properties, and edit a block with read-before-write enforced by passing its exact current content
  • Structured task creation, the only path to a task block, and a narrow cross-namespace channel that changes nothing but a task's status marker
  • An audit line appended to today's journal on every successful write, and blacklisted pages hidden and redacted from every read
Requirements

Logseq running with its HTTP APIs server enabled (Settings → Features, then start it from the plug menu) and an authorization token from those settings, passed as LOGSEQ_API_TOKEN. Installed from PyPI as mcp-server-logseq and launched with uvx over stdio; LOGSEQ_API_URL points at Logseq, default http://localhost:12315. A streamable-http transport is also available for remote clients and requires its own bearer secret, LOGSEQ_MCP_HTTP_TOKEN — the server refuses to start in that mode without one. A config file is optional; the defaults are read-mostly.

Setup effort

One command plus a key — uvx mcp-server-logseq, then supply credentials