skills.rest
bastani-inc
bastani-incOFFICIAL
Explore all skills by this author ā†’
WebsiteGitHub

research-codebase

Document codebase behavior by gathering evidence from source files and research artifacts.

484|76|Updated Oct 23, 2025
Search Tags:#architecture#documentation#compatibility#analysis#synthesis#codebase research#research notes
One-click install
npx skills add https://github.com/bastani-inc/atomic --skill research-codebase-bastani-inc
Or copy as Structured Prompt for Agentā–¼
Please help me install this Agent Skill.
Skill: research-codebase
Source: https://github.com/bastani-inc/atomic/tree/main/packages/workflows/skills/research-codebase
Command: npx skills add https://github.com/bastani-inc/atomic --skill research-codebase-bastani-inc

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you investigate a codebase thoroughly and turn your findings into a structured research document with concrete evidence, file references, and historical context.

Core Features & Use Cases

  • Parallel Codebase Research: Breaks a question into focused subtopics and documents what exists across source files and related research artifacts.
  • Compatibility-Aware Analysis: Captures whether backward compatibility matters so findings reflect the correct downstream constraints.
  • Structured Synthesis: Produces a complete markdown research writeup with summary, detailed findings, code references, architecture notes, and open questions.
  • Use Case: Use it when you need to understand how authentication, data flow, workflows, or another subsystem currently works before making changes.

Quick Start

Ask the skill to research a specific codebase topic and it will gather the relevant files, synthesize the findings, and generate a research document.

Frequently Asked Questions about research-codebase

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

FAQPage Schema
How do I document a codebase architecture with file references and historical context?ā–¼

To document codebase architecture, this skill gathers evidence from repository files and research notes to produce a synthesized markdown document. It includes detailed findings, line-level code references, architecture notes, and open questions.

What is the best way to trace how a specific subsystem currently works before making changes?ā–¼

The best way to trace a subsystem is by breaking the investigation into focused subtopics and reviewing source files in parallel. This captures the current data flow, compatibility constraints, and historical context before modifications.

Can I research backward compatibility constraints across live source files and research notes?ā–¼

Yes, you can research backward compatibility constraints by tracking the compatibility posture across live source files and related research artifacts. This ensures findings reflect the correct downstream constraints for the codebase.

How do I generate a research writeup with frontmatter, code references, and open questions?ā–¼

You generate a research writeup by gathering evidence from the repository and synthesizing the findings into a complete markdown document. The output includes frontmatter, a summary, detailed findings, line-level code references, and open questions.

Does codebase research require parallel source review for component tracing?ā–¼

Yes, codebase research requires parallel source review to effectively trace components and answer architecture questions. Reviewing multiple source files concurrently ensures comprehensive evidence gathering and accurate synthesis.

Related Skills

Explore More Education & Research Skillsā†’
reactjsreactjs
37

react-expert

Research React APIs from local repository sources and save evidence-backed documents.

Official
Advanced
Jeff-StapletonJeff-Stapleton
1

spec-generator

Generate structured markdown specification documents for software projects.

Community
Intermediate
alangebalangeb

git-advanced

Perform Git bisecting, cherry-picking, and log analysis operations locally.

Community
Advanced
paimlpaiml
161

Deep Context Generation with PMAT

Generate compressed LLM-optimized codebase context with PMAT.

Official
Advanced