Verified

Logseq MCP Server

LogSeq knowledge base access for Claude to read, create, and manage your notes.

Data last scanned 2 days ago · Reviewed 2 months ago

20 2 dangerous tools19 tools★ 327PythonMIT

Overview

This server bridges your LogSeq graphs with AI, enabling Claude to perform semantic searches, organize knowledge, and generate documentation. By connecting through the native LogSeq HTTP API, it facilitates complex workflows like summarizing meeting notes, creating study roadmaps, or cross-referencing ideas across your entire database. The tool supports both standard page management and advanced DB-mode graph interaction. Users can leverage optional semantic vector search to find content based on intent and context, rather than simple keyword matching, using local or remote embedding services.

Our verdict

With a security score of 45/100, this server carries a high risk level. The score reflects that the tool is community-maintained and has not yet undergone editorial or security reviews, meaning its authentication and read-only implementations remain unverified. It is suitable for enthusiasts and hobbyists who are comfortable managing their own data security and API token exposure, but it should be approached with caution in environments containing sensitive information.

Zero direct dependencies reduces the supply-chain surface area
Supports semantic vector search via Ollama or OpenAI
Provides functionality for DB-mode graph reading and writing

Community-maintained with no official vendor endorsement
Security architecture remains unreviewed by experts
Requires manual API token management and HTTP server configuration

Tools

Tool	Description	Risk
list_pages	Browse your graph	low
get_page_content	Read page content	low
create_page	Add new pages with structured blocks	medium
update_page	Modify pages (append/replace modes)	medium
delete_page	Remove pages	high
delete_block	Remove a block by UUID	high
update_block	Edit block content by UUID	medium
search	Find content across graph	low
query	Execute Logseq DSL queries	low
find_pages_by_property	Search pages by property	low

Compatibility

Client	Local	Docker	Remote	Read-only
Claude Desktop
Cursor
VS Code
Windsurf
ChatGPT

Frequently asked questions

›What is required to connect the server to Logseq?

You must enable the HTTP API server in Logseq's settings under Features, then start the server and generate an authorization token within the API panel.

›Is the semantic vector search feature required?

No, semantic vector search is optional and requires additional configuration for embeddings via Ollama or an OpenAI-compatible endpoint.

›Does this server support Logseq's DB-mode?

Yes, it supports DB-mode, allowing you to read and write class properties on blocks using specific tools like set_block_properties.

›How are markdown inputs handled by the server?

The server automatically converts markdown into Logseq's native block structure when using the create_page and update_page tools.

›Where is the vector database stored?

The server uses LanceDB, which remains stored locally regardless of whether you choose to use a local Ollama model or a hosted OpenAI-compatible embedding provider.