code-map

Build a cached JSON symbol and call-graph map using Language Server Protocol.

10|6|Updated Feb 8, 2026
One-click install
npx skills add https://github.com/juxt/claude-plugins --skill code-map-juxt
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: code-map
Source: https://github.com/juxt/claude-plugins/tree/main/plugins/code-map/skills/code-map
Command: npx skills add https://github.com/juxt/claude-plugins --skill code-map-juxt

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Navigating unfamiliar codebases, tracing function call relationships, and assessing the impact of code changes is extremely tedious using only grep and manual file reads. This Skill solves that by building a precise, LSP-powered symbol and call-graph map of your project that lets you navigate code by definition and relationship instead of blind text search.

Core Features & Use Cases

  • LSP-Powered Code Mapping: Builds a cached JSON map of project symbols and call edges stored in the .codemap/ directory at the project root, supporting Python, Java, TypeScript, Kotlin, and Clojure projects via language-specific adapters.
  • Targeted Querying: Answer specific questions like "Where is this function defined?", "What calls this function?", or "What breaks if I edit this line?" without rebuilding the entire map.
  • Incremental Refresh: Only rebuilds symbols and edges for files changed since the last map build, saving time on large projects.
  • Use Case: When planning to modify a core payment processing function, use this Skill to trace all callers of the function across the project to identify potential breaking changes before writing any code.

Quick Start

Use the code-map skill to generate a symbol and call-graph map of your current project to trace all code that calls the process_payment function before making edits to it.

Frequently Asked Questions about code-map

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I trace function call relationships in an unfamiliar codebase without using grep?

You can trace function call relationships by building a cached symbol and call-graph map using LSP, which enables precise code navigation by definition and relationship instead of blind text search.

What is the best way to assess the blast radius of proposed code changes across a project?

Assess the blast radius of code changes by querying a cached symbol and call-graph map to identify all callers of a specific function across Python, Java, TypeScript, Kotlin, and Clojure projects before writing any code.

Does code mapping work with Python, Java, TypeScript, Kotlin, and Clojure projects?

Yes, code mapping works with Python, Java, TypeScript, Kotlin, and Clojure projects by utilizing language-specific LSP adapters to extract symbols and call edges for the cached map.

How do I locate symbol definitions and find what calls a specific function in a large project?

Locate symbol definitions and find callers by running targeted queries against the cached JSON map stored in the .codemap/ directory, answering specific relationship questions without rebuilding the entire map.

Can I incrementally refresh a code symbol map to only process changed files?

Yes, you can incrementally refresh the code symbol map to only rebuild symbols and edges for files changed since the last map build, saving time on large projects.

Why use LSP for codebase exploration instead of manual file reads and text search?

LSP provides precise code intelligence for codebase exploration by mapping structural definitions and call relationships, overcoming the limitations of blind text search and manual file reads when navigating unfamiliar code.