cartographer

Map large codebases by orchestrating parallel subagents into CODEBASE_MAP.md and CLAUDE.md.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/rawhit-r/ai-files --skill cartographer-rawhit-r
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: cartographer
Source: https://github.com/rawhit-r/ai-files/tree/main/.gemini/cartographer
Command: npx skills add https://github.com/rawhit-r/ai-files --skill cartographer-rawhit-r

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires tiktoken, and includes scripts (resource) components.

What problem does it solve?

Cartographer automates the creation of CODEBASE_MAP.md and CLAUDE.md by orchestrating parallel subagents to map large codebases, producing architecture, file purposes, dependencies, and navigation guides.

Core Features & Use Cases

  • Maps architecture and file-level details for large projects
  • Generates docs/CODEBASE_MAP.md and updates CLAUDE.md with summaries
  • Supports onboarding to new projects and ongoing maintenance

Quick Start

Run Cartographer to map the current project and generate or update the codebase map.

Frequently Asked Questions about cartographer

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

FAQPage Schema
How do I automatically map and document a large codebase?

You can map a large codebase by running Cartographer, which orchestrates parallel Sonnet subagents to scan files and generate docs/CODEBASE_MAP.md and CLAUDE.md summaries. This automated documentation covers architecture, file purposes, and dependencies across multi-language repositories.

What is the best way to generate architecture documentation for onboarding to a new project?

Generating architecture documentation for project onboarding is best handled by Cartographer, which coordinates parallel subagents to produce a codebase map. It outputs navigation guides and file-level details into docs/CODEBASE_MAP.md and updates CLAUDE.md automatically.

Can I document multi-language repositories within a defined token budget?

Yes, you can document multi-language repositories within a defined token budget using Cartographer. It orchestrates parallel subagents to scan the codebase and generate architecture summaries while respecting the token limits specified in the SKILL.md configuration.

Do I need a SKILL.md frontmatter to map my codebase?

Yes, you need a SKILL.md frontmatter with a name and description to map your codebase. Cartographer requires this file to define the mapping task and orchestrate the parallel subagents that generate the documentation output.

Does Cartographer require tiktoken to scan and document file purposes?

Yes, Cartographer requires the tiktoken dependency to scan and document file purposes. It uses tiktoken for token counting while orchestrating subagents within the defined token budget to map the architecture and generate the codebase documentation.

When should I not use parallel subagents for codebase documentation?

You should not use parallel subagents for codebase documentation on very small projects where manual mapping is faster, or when lacking the required tiktoken dependency and SKILL.md frontmatter. Cartographer is designed for large, multi-language repositories needing architecture reviews and ongoing maintenance.