MCP Catalogs
← Homelumen screenshot

lumen

by ory·★ 190·Score 51

Local semantic code search engine that reduces AI agent token usage by up to 66% through efficient code indexing.

developer-toolsai-llmsearch
21
Forks
10
Open issues
this month
Last commit
2d ago
Indexed

Overview

Lumen is a local semantic code search engine that serves as an MCP server for AI coding assistants. It reduces token usage and processing time by indexing code semantically and providing search capabilities instead of reading entire files. The tool works with Ollama or LM Studio for embeddings and stores everything locally in SQLite. It supports 12 programming languages and provides measurable improvements in performance across multiple benchmarks.

Try asking AI

After installing, here are 6 things you can ask your AI assistant:

you:Enhancing code search for Claude Code, Cursor, Codex and OpenCode AI assistants
you:Reducing AI processing costs for code analysis in large codebases
you:Improving development efficiency by providing semantic code search
you:What programming languages does Lumen support?
you:How does Lumen reduce token usage?
you:Does Lumen require cloud services?

When to choose this

Choose Lumen when working with AI coding assistants and need to significantly reduce token usage and costs while maintaining code quality, especially for TypeScript/JavaScript codebases.

When NOT to choose this

Avoid Lumen if you need to work with closed-source embedding models or require cloud-based vector search capabilities.

Tools this server exposes

3 tools extracted from the README
  • semantic_search

    Perform semantic search on code using vector embeddings

  • health_check

    Check the health status of the Lumen server

  • index_status

    Get the status of the code index

Comparable tools

semantic-searchcode-searchgithub-copilottabnine

Installation

Installation

Claude Code

/plugin marketplace add ory/claude-plugins
/plugin install lumen@ory

Verify by starting a new Claude session and running /lumen:doctor.

Cursor

Lumen ships a native Cursor plugin bundle. Use Cursor's plugin installation workflow with the bundle in .cursor-plugin/ directory.

Codex

codecodex mcp add lumen -- "${CODEX_HOME:-$HOME/.codex}/lumen/scripts/run" stdio

OpenCode

Add @ory/lumen-opencode to the plugin array in your opencode.json:

{
  "plugin": ["@ory/lumen-opencode"]
}

FAQ

What programming languages does Lumen support?
Lumen supports 12 language families including Go, Python, TypeScript, JavaScript, Rust, Ruby, Java, PHP, C/C++, C#, and Dart with semantic chunking.
How does Lumen reduce token usage?
Instead of having AI agents read entire files, Lumen provides semantic search capabilities to find relevant code chunks, reducing the amount of code that needs to be processed.
Does Lumen require cloud services?
No, Lumen is 100% local. It uses Ollama or LM Studio for embeddings and SQLite for storage, with no data leaving your machine.

On Hacker News

Recent discussion from the developer community.

Compare lumen with

lumen vs ultimate_mcp_serverlumen vs mcp-server-chartlumen vs everythinglumen vs filesystemlumen vs time
GitHub →

Last updated · Auto-generated from public README + GitHub signals.