What problem does it solve? End-user guides often ship with coverage holes, mistyped documentation modes, inaccurate steps, or jargon that blocks non-technical readers, and self-review by the author rarely catches these gaps. This Skill acts as an independent acceptance gate that judges a finished guide against a fixed 12-condition bar and returns a machine-parseable verdict with actionable findings. ## Core Features & Use Cases - 12-condition usability and accuracy checklist: Judges goal coverage, Diataxis mode typing, reference completeness, step accuracy with exact UI labels, procedure mechanics, troubleshooting, plain language, accessibility, and findability. - Machine-parseable verdict output: Emits exactly VERDICT: approve or VERDICT: revise plus findings that name the failed condition, location, and concrete fix. - Amend-mode review: Judges versioned deltas on published guides with a staleness sweep across all modes when handed an existing guide plus changed upstreams. - Use Case: After authoring a user guide from a feature-spec and user-flows, run this Skill as the review gate; it flags a missing how-to for the "Reset password" goal and an unexplained acronym, returning findings the author resolves in one revision pass. ## Quick Start Use reviewing-user-guide on the finished end-user guide at docs/guide.md and give me the verdict with actionable findings.
