bmad-code-review

Invoke the review with a target such as a branch diff, commit range, staged changes, or a pasted diff. The workflow gathers the diff, runs parallel review layers, triages findings by severity, and lets you apply patches or defer items.

It runs four layers: Blind Hunter for general issues, Edge Case Hunter for unhandled paths, Verification Gap for untested behavior changes, and Acceptance Auditor for spec compliance. Layers can be customized or disabled through TOML overrides.

Yes, declare no-spec mode explicitly or choose to continue without a spec when asked. In no-spec mode the Acceptance Auditor layer is skipped and ambiguous findings are reclassified as patch or defer instead of decision-needed.

Yes, when a story key is known it updates the story file status to done or in-progress based on review outcome and syncs the matching entry in sprint-status.yaml. Without a story key, only findings are presented.

Failed or empty layers are recorded in a failed-layers list and the review continues with remaining layers. If all findings are dismissed but layers failed, the workflow warns that the review may be incomplete rather than declaring it clean.