Which models does it use?

It typically uses Claude 3 Opus for orchestration and planning, while delegating the file reading and analysis tasks to efficient Sonnet subagents.

How does it handle very large codebases?

It splits the codebase into manageable chunks based on token counts and assigns them to parallel Sonnet subagents to stay within context limits while ensuring comprehensive coverage.

What documentation does it generate?

It creates a detailed CODEBASE_MAP.md file including Mermaid architectural diagrams, module guides, file purposes, and data flow visualizations.

Does it support incremental updates?

Yes, Cartographer detects changes since the last mapping session using git logs or file hashes and only updates the relevant sections of your documentation.

What is Cartographer?

Cartographer is a Claude Code skill that uses parallel AI subagents to scan, map, and document codebases of any size, providing a clear overview of the system architecture.

Cartographer Codebase Mapper

Name: Cartographer Codebase Mapper
Author: kingbootoshi

bykingbootoshi

•

401

Learning & Documentation

Maps and documents complex codebases using parallel AI subagents to provide architectural insights and navigation guides.

About

Cartographer is a specialized Claude Code skill designed to demystify large or unfamiliar repositories by orchestrating parallel Sonnet subagents to analyze directory structures, dependencies, and code patterns. It generates a comprehensive CODEBASE_MAP.md featuring Mermaid architecture diagrams, module guides, and data flow visualizations, while also keeping CLAUDE.md updated with high-level summaries. This skill is essential for onboarding to new projects, maintaining architectural documentation, and ensuring developers can quickly navigate complex file systems without exceeding model token limits.

Key Features

401 GitHub stars
Generates CODEBASE_MAP.md with Mermaid architecture and sequence diagrams
Provides file-level token counting to optimize AI context window usage
Creates detailed navigation guides for common development tasks and module relationships
Orchestrates parallel subagents for high-speed, large-scale codebase analysis
Automatically detects changes to perform incremental documentation updates

Use Cases

Onboarding to a large, undocumented project for the first time
Identifying dependencies and data flows before refactoring complex modules
Generating architectural documentation and diagrams for team members

About

Key Features

401 GitHub stars
Generates CODEBASE_MAP.md with Mermaid architecture and sequence diagrams
Provides file-level token counting to optimize AI context window usage
Creates detailed navigation guides for common development tasks and module relationships
Orchestrates parallel subagents for high-speed, large-scale codebase analysis
Automatically detects changes to perform incremental documentation updates

Use Cases

Onboarding to a large, undocumented project for the first time
Identifying dependencies and data flows before refactoring complex modules
Generating architectural documentation and diagrams for team members