What problem does it solve? Teams often plan perfect solutions to the wrong problem because the observation and the assumed cause get treated as one fact. This Skill separates what is actually observed from what is assumed before any planning begins, preventing wasted effort on misframed tasks. ## Core Features & Use Cases - Framing Separation: Captures the reported observation, the user's stated cause, and the proposed direction as three distinct items so assumptions stay visible. - Hypothesis Investigation: Builds a dimension map of where the problem could originate, then dispatches parallel read-only sub-agents to gather file:line evidence for each hypothesis. - Frame Brief Output: Produces a structured frame.md artifact with a hypothesis evidence table, confidence rating, and a reframed or confirmed problem statement ready to hand off to planning. - Use Case: A user says "the API is broken, let's rewrite the handler." This Skill investigates whether the observation actually originates in the handler, the input validation, or elsewhere, and confirms or reframes the problem before a plan is written. ## Quick Start Invoke the frame skill with a description of the bug or scope question you are about to plan, and let it verify the framing before you commit to a solution.
