#192 · Primary category: AI Coding Assistants
jcodemunch-mcp
Cut AI token costs 95%+ on code exploration. The leading MCP server for precise, symbol-level GitHub code retrieval via tree-sitter AST. Works with Claude Code, Cursor & any MCP client. 313B+ tokens saved.
Project last updated:08/29/26
GitHub Stars
2.6K
Forks
359
Contributors
55
License
Other
Why we included this project
Large codebases make agentic editors waste tokens fast, since tools like Claude Code or Cursor will read whole files just to answer a small question. jCodeMunch sidesteps that by indexing your repository locally with tree-sitter and serving agents only the exact definitions they ask for. The maintainers publish a reproducible benchmark putting the savings at roughly 28x fewer tokens than a grep-and-read agent, and the structural queries are what make it worth using day to day: finding what imports a symbol or checking the blast radius before a change is something native tools rarely do well. It installs as a standard MCP server, so your existing setup stays intact, and indexing runs local-first if you would rather not send code to a hosted service.
Articles for this project
No articles for this project yet.
To suggest a topic or contribute an article, contact us.
Related projects in this category
superpowers
An agentic skills framework & software development methodology that works.
ECC
The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
skills
Skills for Real Engineers. Straight from my .agents directory.
opencode
The open source coding agent.
andrej-karpathy-skills
A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.