iterative-plan-review

Validate feature plan assumptions by cross-referencing repository context, codebase patterns, and project documentation to confirm or refute claims. This evidence-driven plan review identifies primary and secondary assumptions to reduce guesswork and justify necessary edits.

Detect overlaps and edge cases by comparing planned items against existing codebase patterns and documentation. This risk analysis surfaces redundancies, gaps, and ambiguities, producing structured review findings and actionable plan edits.

Run a lightweight iteration by providing your plan to surface immediate findings and suggested edits. Apply the evidence-backed changes as needed, or use team-mode for collaborative, multi-agent reviews.

Evidence-driven planning works with ADRs and project documentation by collecting and cross-referencing them to justify each plan change. This ensures deterministic alignment between plan edits and repository artifacts.

Maintain an audit history of plan reviews through companion artifacts, specifically review-findings.md and review-iteration-history.md files. These outputs ensure transparent tracking of structured findings and applied edits over time.

Avoid iterative plan reviews when a codebase lacks sufficient context, patterns, or documentation to validate assumptions. Without repository evidence to cross-reference, the review cannot produce deterministic, actionable plan edits.