j

jcodemunch-mcp

jCodeMunch is an efficient MCP server that uses a tree-based parser (AST) to precisely retrieve GitHub code, significantly reducing AI token costs.

🌍 OverseasFree + PaidOpen sourceDev开源API
Platforms: WebAPI
Region
Overseas
Pricing
Free + Paid
Open source
Yes
GitHub Stars
★ 1.9k
Source
GitHub
Added
2026-06-25
Last verified
2026-06-25
jcodemunch-mcp

Overview

jCodeMunch is a leading MCP server that leverages a tree-based parser (AST) to accurately retrieve GitHub code, substantially lowering AI token costs in code exploration. It supports Claude Code, Cursor, and any MCP-compatible client. By indexing code repositories once, jCodeMunch enables agents to retrieve only the exact code needed, reducing code reading token usage by over 95% in retrieval-intensive workflows. Ideal for developers and teams, it helps enhance efficiency in code exploration and management.

Key features

  • Reduces AI token costs by over 95%
  • Supports multiple MCP clients
  • Precise code retrieval via tree-based parser (AST)
  • Index once, query multiple times

Use cases

Code explorationCode retrievalCode understandingCode structure exploration

Pros

  • Significantly reduces token costs
  • Supports multiple development tools
  • Precise code retrieval
  • Easy to install and configure

Limitations / notes

  • Requires some learning curve
  • Depends on network connectivity

Who it's for

DevelopersEnterprise teamsAI developers

This overview was compiled by AI from public sources and may contain inaccuracies — please refer to the official site.

FAQ

Is it free?

Personal use is free; commercial use requires payment.

Which development tools are supported?

Supports MCP-compatible clients such as Claude Code, Cursor, VS Code, and Codex CLI.

Is Chinese supported?

Documentation is primarily in English; Chinese support is not explicitly mentioned.

Can it be used commercially?

Yes, it can be used commercially, but a commercial license must be purchased.

Similar agents

Something wrong? Let us know on the About page and we'll fix it.