research

Document a codebase as-is with parallel sub-agents and metadata-rich architecture references.

1|Updated Jan 2, 2026
One-click install
npx skills add https://github.com/tfunk1030/vibe --skill research-tfunk1030
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: research
Source: https://github.com/tfunk1030/vibe/tree/main/.claude/skills/research
Command: npx skills add https://github.com/tfunk1030/vibe --skill research-tfunk1030

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides a precise, read-only map of the current codebase, preserving historical context via a thoughts directory. It enables engineers to understand existing structures and interactions without making changes.

Core Features & Use Cases

  • Document the codebase as-is with a thoughts directory for historical context to support onboarding and knowledge transfer.
  • Spawn parallel sub-agents to research relevant components and synthesize findings into a complete technical map.
  • Produce metadata-rich documentation including architecture, references, and historical context, without modifying the code.

Quick Start

Run the research harness to document the codebase and generate a metadata-rich map without altering any code.

Frequently Asked Questions about research

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

FAQPage Schema
How do I document codebase architecture without modifying the source code?

You can document a codebase architecture without modifying source code by running this research harness to generate a metadata-rich technical map. It spawns parallel sub-agents to synthesize findings, producing read-only documentation that includes references and historical context.

What is the best way to preserve historical context and thoughts during codebase research?

The best way to preserve historical context during codebase research is to utilize a thoughts directory. This approach captures architectural metadata and historical decisions, creating a complete technical map that supports onboarding and knowledge transfer without altering code.

How to generate technical documentation for onboarding new engineers to an existing codebase?

To generate technical documentation for onboarding, run a research process that spawns parallel sub-agents to investigate relevant components. These agents synthesize their findings into a comprehensive technical map containing architecture details and historical context.

Does this codebase documentation approach propose changes to the current architecture?

No, this codebase documentation approach does not propose changes. It functions as a precise, read-only map that documents the architecture as-is, strictly focusing on metadata synthesis and historical context without modifying the existing code structure.

Can I use parallel sub-agents to research and synthesize metadata for large software engineering projects?

Yes, you can use parallel sub-agents to research and synthesize metadata for large software engineering projects. This method efficiently maps relevant components and references, compiling a comprehensive technical overview without proposing codebase changes.