What problem does it solve? Raw customer requests arrive as noisy mixes of chat logs, screenshots, workarounds, and wishes. This Skill turns that noise into a verifiable Intake Brief with facts, hypotheses, scope boundaries, and open questions before any code investigation begins. ## Core Features & Use Cases - Input Classification: Routes incoming material as intake, debug (traces/stacks), explore (code investigation), or change requests, with automatic handoff to the right OpenSpec command. - Clarification Loop: Runs a Readiness Check and asks up to 3 targeted questions per iteration (max 3 iterations) until the brief meets quality criteria or the user fixes it as-is. - Structured Intake Brief: Separates verified facts, hypotheses, and noise; defines current/target scenarios, scope boundaries, technical context, and open questions split by resolvable-next and stakeholder-only. - Use Case: A customer sends a long email with screenshots asking to "improve the invoice form". The Skill classifies the input, clarifies the goal and scope through questions, then produces a brief and hands off to /opsx:explore or /opsx:new. ## Quick Start Paste the raw customer request or notes into the chat and ask the assistant to run the openspec-intake skill to produce an Intake Brief.
