jcodemunch-mcp
byjgravelle· 2 MCP servers
The leading, most token-efficient MCP server for precise GitHub source code retrieval via tree-sitter AST parsing. Cut AI token costs 95%+ on code exploration — stop burning your context window reading entire files.
How to install jcodemunch-mcp?
jgravelle/jcodemunch-mcp
$git clone https://github.com/jgravelle/jcodemunch-mcpInstalls into the current project.
›Prefer a prompt? Paste this to your agent
Install jcodemunch-mcp by running `git clone https://github.com/jgravelle/jcodemunch-mcp`, then use it for the current task and follow its documentation at https://github.com/jgravelle/jcodemunch-mcp.
README.md
| 1 | # jCodeMunch MCP |
| 2 | |
| 3 | **The leading, most token-efficient MCP server for precise GitHub source code retrieval via tree-sitter AST parsing.** Cut AI token costs **95%+** on code exploration — stop burning your context window reading entire files. |
| 4 | |
| 5 | > **Real results, live from production** |
| 6 | > **335B+ tokens saved** · **48,000+ developers** · **$1.69M+ in AI spend avoided** · **40,000+ kg CO₂ prevented** |
| 7 | > Live telemetry at **[jcodemunch.com](https://jcodemunch.com/)** — benchmark: **95% average token reduction** (15 tasks / 3 repos, 99.8% peak). |
| 8 | |
| 9 | Works with **Autohand Code**, **Claude Code**, **Cursor**, **VS Code**, **Codex CLI**, **Continue**, **Windsurf**, and any MCP-compatible client. |
| 10 | |
| 11 | --- |
| 12 | |
| 13 | #### One-click installs: |
| 14 | |
| 15 | [](vscode:mcp/install?%7B%22name%22%3A%20%22jcodemunch%22%2C%20%22command%22%3A%20%22uvx%22%2C%20%22args%22%3A%20%5B%22jcodemunch-mcp%22%5D%7D) |
| 16 | [](vscode-insiders:mcp/install?%7B%22name%22%3A%20%22jcodemunch%22%2C%20%22command%22%3A%20%22uvx%22%2C%20%22args%22%3A%20%5B%22jcodemunch-mcp%22%5D%7D) |
| 17 | [](cursor://anysphere.cursor-deeplink/mcp/install?name=jcodemunch&config=eyJjb21tYW5kIjogInV2eCIsICJhcmdzIjogWyJqY29kZW11bmNoLW1jcCJdfQ==) |
| 18 | [](#works-with) |
| 19 | [](#works-with) |
| 20 | |
| 21 | Prefer the command line? |
| 22 | |
| 23 | ``` |
| 24 | pip install jcodemunch-mcp |
| 25 | uvx jcodemunch-mcp |
| 26 | ``` |
| 27 | |
| 28 | For pinned/B2B deployments that want a version-stable install channel independent of PyPI, install straight from the repo (requires `git`, builds from source): |
| 29 | |
| 30 | ``` |
| 31 | pip install git+https://github.com/jgravelle/jcodemunch-mcp.git |
| 32 | uvx --from git+https://github.com/jgravelle/jcodemunch-mcp.git jcodemunch-mcp |
| 33 | ``` |
| 34 | |
| 35 | Quickstart - https://github.com/jgravelle/jcodemunch-mcp/blob/main/QUICKSTART.md |
| 36 | |
| 37 | A crapload of detailed info: http://jcodemunch.com/ |
| 38 | |
| 39 | **Live OSS code-health observatory** — weekly six-axis health snapshots |
| 40 | of Express, FastAPI, Gin, Pydantic, Django, Flask, NestJS, Cobra, and |
| 41 | this very repo: https://jgravelle.github.io/jcodemunch-observatory/ |
| 42 | |
| 43 | **Token Cost Radar** — daily intelligence on AI token costs, minimization |
| 44 | strategies, and budget trends for teams running Claude Code / Cursor / MCP: |
| 45 | https://jcodemunch.com/radar/ |
| 46 | |
| 47 | <!-- mcp-name: io.github.jgravelle/jcodemunch-mcp --> |
| 48 | |
| 49 | |
| 50 | ## FREE FOR PERSONAL USE |
| 51 | **Use it to make money, and Uncle J. gets a taste. Fair enough?** [details](#commercial-licenses) |
| 52 | |
| 53 | Our guarantee: If jCodeMunch doesn't pay for itself, you don't pay for jCodeMunch! |
| 54 | |
| 55 | --- |
| 56 | |
| 57 | |
| 58 | ## Cut code-reading token usage by **95% or more** with precise symbol retrieval |
| 59 | |
| 60 | Most AI agents explore repositories the expensive way: |
| 61 | |
| 62 | open entire files → skim thousands of irrelevant lines → repeat. |
| 63 | |
| 64 | That is not “a little inefficient.” |
| 65 | That is a **token incinerator**. |
| 66 | |
| 67 | **jCodeMunch indexes a codebase once and lets agents retrieve only the exact code they need**: functions, classes, methods, constants, outlines, and tightly scoped context bundles, with byte-level precision. |
| 68 | |
| 69 | In retrieval-heavy workflows, that routinely cuts code-reading token usage by **95%+** because the agent stops brute-reading giant files just to find one useful implementation. |
| 70 | |
| 71 | | Task | Traditional approach | With jCodeMunch | |
| 72 | | ---------------------- | ------------------------- | ------------------------------------------- | |
| 73 | | Find a function | Open and scan large files | Search symbol → fetch e |
Related
6 picksType
- minimax-coding-plan-mcpSpecialized MiniMax Model Context Protocol (MCP) server designed for coding-plan usersFeb 2026
- neoNeo.mjs is a self-evolving software organism: a professional end-to-end AI engineering team whose cross-model swarm inhabits live apps via Neural Link, Active Hybrid GraphRAG, DreamService, and self-healing loops.Jul 2026
- shadcn-ui-mcp-serverA Model Context Protocol (MCP) server for shadcn/ui components, providing AI assistants with access to component source code, demos, blocks, and metadata.May 2026
- kicad-mcp-proProduction-grade MCP server for KiCad EDA—PCB design, DRC, simulation, BOM, DFM, and manufacturing.Jul 2026
- taiga-ui-mcpModel Context Protocol server providing Taiga UI documentation search and scaffolding tools.Jul 2026
- mcpOfficial MCP server for Magic UI.Apr 2026
