pw-research

Coordinate three parallel research agents and validate findings against the codebase.

Updated Apr 14, 2026

Search Tags:#automation #workflow #research #codebase #validation #agents #parallel

Ready to run

npx skills add https://github.com/patrob/skills --skill pw-research

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: pw-research
Source: https://github.com/patrob/skills/tree/main/plugins/claude-skills/skills/pw-research
Command: npx skills add https://github.com/patrob/skills --skill pw-research

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Coordinate deep parallel research to understand a problem from multiple angles and validate findings against the codebase.

Core Features & Use Cases

Launch 3 parallel research agents to explore the problem from web, codebase, and project context perspectives.
Aggregate findings and synthesize a unified research document with executive summary and open questions.
Validate findings against actual code and project context, ensuring correctness before planning.

Quick Start

Provide a research topic and problem statement to initiate a three-agent study.

Frequently Asked Questions about pw-research

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

FAQPage Schema

How do I run parallel research across web sources and my codebase?▼

You can run parallel research by launching three agents to explore web sources, analyze the codebase, and gather project context, then synthesizing the findings into a unified document.

What is the best way to validate research findings against actual code?▼

The best way to validate research findings is to cross-check aggregated results against actual code and project context, which ensures accuracy and consistency before planning.

How does a multi-agent research workflow work for codebase analysis?▼

A multi-agent research workflow coordinates separate agents for web research, codebase analysis, and project context to gather diverse perspectives, then aggregates them into a consolidated document.

Do I need to provide a specific problem statement to start automated research?▼

Yes, you need to provide a research topic and problem statement to initiate the three-agent study and generate a validated research report.

Can I use parallel research agents to prepare for project planning?▼

Yes, you can use parallel research agents to produce a consolidated research document and validation report specifically designed to inform subsequent planning phases.

What limitations exist when validating web research against a local codebase?▼

Validation depends on the accuracy of the codebase context provided; if the project context is incomplete, the validation report may miss inconsistencies between web findings and actual code.