glommer

glommer / codemogger

Public

Codemogger is a code indexing library and MCP server for AI coding agents

176
12
100% credibility
Found Feb 24, 2026 at 53 stars 3x -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

Codemogger is a local library that indexes codebases by parsing files into semantic chunks, embedding them for similarity search, and enabling fast keyword and semantic queries via a single database file.

How It Works

1
📖 Discover codemogger

You hear about codemogger, a handy tool that helps AI assistants quickly understand and explore any codebase without needing online services.

2
🛠️ Add it to your tools

You easily install codemogger on your computer with a single command, making it ready to use right away.

3
📁 Prepare your project

You point codemogger at your project folder, and it scans your files, organizing code pieces like functions and classes into a personal search guide.

4
🔍 Start searching

You ask natural questions like 'how does login work?' or look up specific names, and instantly see the most relevant code spots with previews.

5
🤖 Team up with AI

You connect codemogger to your AI coding helper, letting it navigate your code smartly and give precise answers.

🎉 Unlock fast code discovery

Your AI now finds exactly what you need in seconds across huge projects, making coding faster and more fun.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 53 to 176 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is codemogger?

Codemogger is a TypeScript library and CLI for indexing codebases, turning source files into searchable semantic chunks for AI coding agents. It scans directories, parses code into functions, classes, and structs using tree-sitter, embeds them locally with a lightweight model, and stores everything in a single SQLite file supporting vector and full-text search. Developers get instant semantic queries like "authentication middleware" or precise keyword lookups, all without Docker, servers, or API keys—one command indexes, another searches.

Why is it gaining traction?

It crushes ripgrep on speed (25x-370x faster keyword search) and quality, surfacing exact definitions instead of file floods, while semantic mode nails conceptual queries across massive repos like Kubernetes. The MCP server plugs straight into agents like Claude Code or OpenCode via simple JSON config, and the SDK lets you supply your own embedder for custom setups. Local embeddings mean zero latency or costs, hooking devs tired of cloud RAG hassles.

Who should use this?

AI coding agent builders integrating codebase awareness into tools like custom Claude or OpenCode setups. TypeScript/Node devs needing offline code search in editors or scripts. Teams evaluating local RAG for monorepos, especially Rust/Go/Python shops benchmarking against grep.

Verdict

Grab it for agent prototypes—benchmarks deliver, docs are crisp, CLI/SDK shine—but with 35 stars and 1.0% credibility, it's early; test on your repo before production. Solid foundation for local code indexing.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.