What problem does it solve? Turning a requirement document or a plain-language request into a buildable spec usually requires slow, deep analysis. This Skill provides a fast-lane alternative that bounds the deliverable to exactly the asked scope, grounds every requirement in real code with file:line evidence, and produces a confirmed report plus derived spec without deep-mode sweeps or scout subagents. ## Core Features & Use Cases - Doc-optional intake: Accepts a PDF by path, pasted document text, or a plain-language request, auto-detecting prose, audit, or requirement mode. - Evidence-or-mark grounding: Maps each requirement to real files with quote-anchored file:line evidence, or tags it ASSUMPTION/UNVERIFIED and turns it into a user question. - Scope bounding with challenges: Keeps the deliverable to exactly the asked scope, challenges ambiguities with recommended options, and derives requirement-spec.md only after the user confirms the report. - Use Case: You ask to add a "resend invite" action with no document. The Skill maps the request to invites/service.py and invites/routes.py, asks one blocking question about token expiry with a recommended option, then writes report.md and requirement-spec.md ready for handoff to the build lane. ## Quick Start Ask the assistant to run /orc-analyze-mini on your requirement, for example: quickly analyze this request to add a resend invite action and produce a scoped spec.
