What problem does it solve? Complex research tasks that require investigating many sources can flood the main conversation with redundant fetches and wasted tokens. This Skill defines when and how to delegate such investigations to a research subagent, and how to trust its results without re-querying the same sources. ## Core Features & Use Cases - Delegation Thresholds: Clear criteria for when to delegate (3+ sources, synthesis, comprehensive coverage) versus when to answer directly with a single lookup. - Subagent Result Authority: Rules that treat subagent output as authoritative, preventing redundant re-fetching of URLs, repositories, and documentation. - Prompt Structuring Guidance: A template for writing self-contained research prompts with objectives, scope, sources, deliverables, and success criteria. - Use Case: When evaluating which authentication library to adopt, delegate a research agent to compare official docs, GitHub repos, and community discussions, then receive a comparison table and recommendation without re-checking any source yourself. ## Quick Start Ask the assistant to research and compare three logging frameworks for your project by delegating to a research agent with defined sources and deliverables.
