Laps/Labs

Resources / skill

Research Skill

Deep context gathering before planning — launches parallel agents to search docs, web and codebase, then synthesizes findings into actionable context.

Use with your AI agent
Fetch and apply this skill from Laps Labs:
https://lapslabs.com/resources/skills/research.md

Use it as the base for "Research Skill". Adapt to this project's stack and conventions.

Or fetch the raw Markdown directly: /resources/skills/research.md

Deep context gathering before any planning or implementation. Sits between scope alignment and plan mode (how to build it).

How to research

Step 1: Clarify before you research (mandatory — never skip)

Before reading a single file or launching any agent, ask clarifying questions. Identify every place where there are 2+ plausible interpretations — scope, intent, constraints, approach, priority — and ask about those specifically.

Step 2: Launch parallel research agents

Fan out independent search agents across docs, the web, and the codebase. Each returns a focused summary rather than raw dumps.

Step 3: Synthesize

Merge findings into a single actionable context document the planning step can consume directly.

Part of the Laps Labs agent resource library. Fetch the raw Markdown for your own agent at the .md URL of this page.