What problem does it solve? Reviewing a code change without structure leads to missed risks and shallow approvals. This Skill turns an LLM into a guided review companion that walks a human through a change — from intent and orientation, through a concern-based walkthrough and risk-spot detail pass, to hands-on testing suggestions and a final approve/rework/discuss decision. ## Core Features & Use Cases - Structured 5-Step Review Flow: Orientation (intent + diff stats), Walkthrough (concern-grouped path:line stops), Detail Pass (risk-tagged hotspots like auth, schema, billing), Testing (observable behavior suggestions), and Wrap-Up (decision prompt). - Review Trail Generation: When no author-produced Suggested Review Order exists, it generates one from the diff by reading changed files and grouping cohesive design concerns. - Customizable Workflow: A customize.toml file supports team and user overrides for activation steps, persistent facts, and post-completion behavior, merged via BMad structural rules. - Use Case: After finishing a feature branch, say "walk me through this change" — the Skill identifies the change from conversation context, sprint status, or git state, computes surface-area stats, and guides you to a confident ship/rework decision. ## Quick Start Ask the assistant to checkpoint or walk you through the current branch or PR you want to review.
