research

Verify technical assumptions through iterative web research and codebase exploration.

2|Updated Jan 14, 2026

Search Tags:#automation #planning #documentation #research #codebase #verification

Ready to run

npx skills add https://github.com/nrempel/claude-research --skill research-nrempel

💬 Prompt for ChatGPT / Claude Chat▼

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

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill solves the problem of unreliable implementation plans caused by unverified assumptions, ensuring that every technical decision is backed by evidence.

Core Features & Use Cases

Systematic Verification: Automatically flags and verifies assumptions against documentation, codebase patterns, and external APIs.
Stateful Research Sessions: Maintains a persistent research log across context compactions, allowing for complex, multi-step investigations.
Parallel Agent Execution: Launches concurrent sub-agents to perform web searches and codebase exploration simultaneously.

Quick Start

Invoke the research skill by typing /research followed by the task you need to verify.

Frequently Asked Questions about research

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

FAQPage Schema

How do I verify technical assumptions before starting a software implementation?▼

You can flag technical assumptions for automatic verification against documentation, codebase patterns, and external APIs. This process grounds implementation plans in documented evidence rather than unverified hypotheses, ensuring reliable project planning.

How does stateful research maintain a verified fact log across context compactions?▼

Stateful research sessions maintain a persistent log across context compactions to preserve verified facts and refuted hypotheses during complex investigations. This ensures iterative web research and codebase exploration remain coherent throughout multi-step project planning.

What is the best way to explore a codebase and search the web simultaneously?▼

Parallel agent execution launches concurrent sub-agents to perform web searches and codebase exploration simultaneously. This accelerates systematic technical assumption verification by gathering documented evidence from multiple independent sources at once.

Can I integrate systematic research directly into my development workflow?▼

Yes, systematic research integrates directly with development workflows to ensure implementation plans are grounded in documented evidence. It identifies and validates technical assumptions through iterative web research and codebase exploration during active project planning.

Does automated assumption verification work without external dependencies?▼

Yes, automated assumption verification requires no external dependencies. It systematically validates technical assumptions by cross-referencing documentation, codebase patterns, and external APIs directly within your existing development workflow and planning environment.