What problem does it solve? Vague research requests like "look into X" force research agents to guess scope, sources, and completion criteria, producing unfocused reports that waste long-running agent runs. This Skill sharpens the ask into a tight, sourceable brief before any subagent is dispatched. ## Core Features & Use Cases - Five-slot brief structure: Fills question, scope boundary, sources, output shape, and done-criteria before dispatching any research work. - Subagent routing guidance: Directs the brief to the right agent type — research for citation-heavy external work, explore for codebase lookups, task for command-output checks. - Artifact placement rules: Decides where the resulting report lives (project docs vs. cross-project research vault) before the agent starts, preventing orphan reports. - Use Case: Before asking a research agent "which observability stack should we use", run this Skill to produce a brief specifying the cardinality constraints, candidate tools, sources to check, and what a sufficient answer looks like. ## Quick Start Ask the assistant to turn your rough research question into a research brief before launching a research subagent.
