skills.rest
atomicinnovation
atomicinnovationOFFICIAL
Explore all skills by this author →
GitHub

research-codebase

Analyze codebases and generate structured research documents with YAML frontmatter.

30|1|Updated Mar 15, 2026
Search Tags:#parallel-agents#repository-analysis#codebase-research#file-references#research-document#github-permalinks
One-click install
npx skills add https://github.com/atomicinnovation/accelerator --skill research-codebase-atomicinnovation
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: research-codebase
Source: https://github.com/atomicinnovation/accelerator/tree/main/skills/research/research-codebase
Command: npx skills add https://github.com/atomicinnovation/accelerator --skill research-codebase-atomicinnovation

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill helps developers and reviewers quickly understand unknown or complex codebases by locating relevant files, extracting evidence, and synthesising a concise, actionable research document that preserves file references and context without overloading the main conversation.

Core Features & Use Cases

  • Parallel sub-agent orchestration: Spawns specialised read-only agents to locate files, analyse code, find patterns, and examine documentation concurrently.
  • Evidence-first research output: Produces a timestamped research document with YAML frontmatter, file paths, line references, metadata, and optional GitHub permalinks for reliable traceability.
  • Use Case: Triage a ticket asking "How does authentication work?" by discovering related files, summarising workflows, and producing a developer-ready research brief in meta/research.

Quick Start

Run the research-codebase skill and provide a focused question such as "How does authentication work in this repository?" to generate a structured research document in meta/research.

Frequently Asked Questions about research-codebase

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

FAQPage Schema
How do I analyze an unfamiliar codebase to understand how a specific feature works?▼

To analyze an unfamiliar codebase, you can conduct comprehensive repository research that locates relevant files, extracts evidence, and synthesizes findings into a structured report with file references and GitHub permalinks.

What is the best way to research a large repository without overloading the main conversation context?▼

The best way to research a large repository is using parallel sub-agent orchestration, which spawns read-only agents to analyze code concurrently and outputs a concise YAML-frontmatter research document in a meta/research directory.

Can I map implementation details and component relationships across an unknown codebase?▼

Yes, you can map implementation details and component relationships by orchestrating parallel sub-agents to locate files, analyze patterns, examine documentation, and gather metadata into a single actionable research brief.

How to generate a traceable research document with file paths and Git permalinks for a codebase review?▼

Generate a traceable research document by running read-only file reads across the codebase, gathering metadata and Git permalinks, and outputting a timestamped YAML-frontmatter file to preserve file references and context.

Does codebase research with parallel agents modify my repository files?▼

No, codebase research with parallel agents performs read-only file reads exclusively, ensuring your repository files remain unmodified while orchestrating sub-agents to gather evidence and synthesize findings.

When should I use parallel agents for repository analysis instead of manual code exploration?▼

Use parallel agents for repository analysis when triaging tickets or navigating large, unfamiliar codebases where you need targeted technical answers, component mappings, and implementation details synthesized quickly.

Related Skills

Explore More Software Engineering Skills→
VMAFxVMAFx
3

dev-llm-review

Run a local LLM code review on a single file with vmaf-dev-llm.

Official
Intermediate
HaydernCenterpointHaydernCenterpoint

github-triage

Generate read-only triage reports for GitHub issues and PRs.

Community
Intermediate
j-showj-show

clean-helper

Identify and remove dead code and deprecated branches via static analysis.

Official
Advanced
massgenmassgen
1.1k

file-search

Locate text and structural code patterns across large codebases using ripgrep and ast-grep.

Official
Basic