gsd-map-codebase

Coordinate parallel agents to generate seven codebase documents in .planning/codebase/.

Updated Jan 25, 2026
One-click install
npx skills add https://github.com/MINT-IA/MINT --skill gsd-map-codebase-mint-ia
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: gsd-map-codebase
Source: https://github.com/MINT-IA/MINT/tree/main/.claude/skills/gsd-map-codebase
Command: npx skills add https://github.com/MINT-IA/MINT --skill gsd-map-codebase-mint-ia

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Analyze codebases by coordinating parallel gsd-codebase-mapper agents to produce structured documentation in the .planning/codebase/ folder, enabling faster onboarding and risk reduction during maintenance.

Core Features & Use Cases

  • Parallel mapper agents explore distinct focus areas and write dedicated documents directly to the .planning/codebase/ directory.
  • Generates a 7-document codebase map (e.g., STACK.md, INTEGRATIONS.md, ARCHITECTURE.md, STRUCTURE.md, CONVENTIONS.md, TESTING.md, CONCERNS.md) to capture current state and design rationale.
  • Useful for brownfield projects, onboarding to unfamiliar repositories, and before major refactors.
  • Supports incremental updates by re-running the mapper workflow to refresh the codebase map.

Quick Start

Run the /gsd-map-codebase command to initialize the codebase mapping and produce the .planning/codebase/ documents.

Frequently Asked Questions about gsd-map-codebase

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

FAQPage Schema
How do I generate codebase architecture documentation for onboarding to an unfamiliar repository?

Codebase architecture documentation is generated by running the /gsd-map-codebase command, which coordinates parallel agents to map subsystems and write seven structured documents directly to the .planning/codebase/ directory for onboarding.

What is the best way to map a brownfield codebase before a major refactor?

Mapping a brownfield codebase before a major refactor is best done by coordinating parallel mapper agents to produce a persistent, up-to-date map of architecture, structure, tests, and concerns in .planning/codebase/, ensuring deterministic outputs for subsequent planning steps.

How does parallel agent mapping work for codebase discovery?

Parallel agent mapping works by an orchestrator coordinating four parallel mapper agents, each focusing on a distinct subsystem to write seven required documents directly to the planning folder, while the orchestrator only collects confirmations for deterministic outputs.

Can I refresh codebase documentation incrementally as an evolving project changes?

You can refresh codebase documentation incrementally by re-running the mapper workflow to update the existing codebase map in .planning/codebase/, ensuring the architecture and structure documents reflect the latest repository state without starting from scratch.

What documents are produced when mapping codebase architecture and conventions?

Mapping codebase architecture and conventions produces a seven-document set in .planning/codebase/, including STACK.md, INTEGRATIONS.md, ARCHITECTURE.md, STRUCTURE.md, CONVENTIONS.md, TESTING.md, and CONCERNS.md to capture the repository's current state.

Do I need external data sources to map codebase structure and conventions?

No external data sources are needed to map codebase structure and conventions; the mapping requires only the repository state itself, orchestrating parallel agents to generate the required planning documents deterministically.