What problem does it solve? When a task needs multiple analytical lenses, it is easy to overload the context with every plausible skill or to blur authority boundaries. This Skill helps you choose the smallest useful skill set with explicit limits on fan-out, authority, and stop conditions. ## Core Features & Use Cases - Bounded Skill Selection: Picks one primary lens, one supporting lens, and at most one verifier unless the task demonstrates a larger need. - Explicit Composition Contract: Returns selected lenses, the context each receives, authority, fan-out limit, execution order, evidence, and stop conditions. - Provider-Neutral Recipes: Recipes select skills only, never providers, models, or authority, keeping runtime requests within closed semantic actions. - Use Case: When planning a design review, use this Skill to combine a brainstorm lens with a code-review verifier, defining exactly what context each receives and when to stop. ## Quick Start Ask the agent to compose skills for your task, for example by saying which skills should I use to review this architecture change with an independent verifier.
