researching

Decompose codebase questions into sub-questions for parallel locator, analyzer, and pattern-finder agents.

1|Updated Jan 31, 2026
One-click install
npx skills add https://github.com/jugrajsingh/skillgarden --skill researching-jugrajsingh
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: researching
Source: https://github.com/jugrajsingh/skillgarden/tree/main/plugins/researcher/skills/researching
Command: npx skills add https://github.com/jugrajsingh/skillgarden --skill researching-jugrajsingh

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill automates in-depth codebase research by decomposing complex questions, dispatching specialized AI agents in parallel, and synthesizing their findings into a structured, citable report.

Core Features & Use Cases

  • Decomposition: Breaks down broad research questions into specific, actionable sub-questions.
  • Parallel Agent Dispatch: Utilizes 'locator', 'analyzer', and 'pattern-finder' agents concurrently for efficient information gathering.
  • Persistent Reporting: Generates detailed reports with file:line citations, ensuring verifiability and aiding future reference.
  • Use Case: When investigating a new feature's implementation, use this Skill to pinpoint relevant files, understand data flow, and identify existing patterns, all synthesized into a single, easy-to-understand document.

Quick Start

Use the researching skill to find out how authentication works in the project.

Frequently Asked Questions about researching

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

FAQPage Schema
How do I conduct parallel codebase research to understand a new feature's implementation?

Parallel codebase research decomposes broad questions into sub-questions, dispatching locator, analyzer, and pattern-finder agents concurrently to pinpoint files, understand data flow, and synthesize findings into a single document.

What is the best way to generate citable documentation for complex codebase questions?

Generating citable documentation involves synthesizing parallel agent findings into a persistent report with file:line references, ensuring verifiability and aiding future reference for complex codebase questions.

Can I use multi-agent research to analyze data flow and identify existing patterns in my project?

Yes, multi-agent research uses locator, analyzer, and pattern-finder agents concurrently to analyze data flow and identify existing patterns, synthesizing the findings for comprehensive understanding.

How does codebase analysis work when investigating unfamiliar code structure?

Codebase analysis works by breaking down broad research questions into specific sub-questions, dispatching specialized AI agents in parallel to gather information, and synthesizing their findings into a structured report.

Do I need any external dependencies to run parallel research agents on my codebase?

No external dependencies are required. The parallel research process operates independently to decompose queries, dispatch agents, and generate detailed reports with file:line citations.

When should I not use a multi-agent approach for codebase analysis?

You should avoid multi-agent codebase analysis for simple, single-file queries where the overhead of decomposing questions and dispatching parallel agents outweighs the benefit of synthesized reporting.