[IMPORTANT] Use TaskCreate to break ALL work into small tasks BEFORE starting — including tasks for each file read. This prevents context loss from long files. For simple tasks, AI MUST ask user whether to skip.

Evidence Gate: MANDATORY IMPORTANT MUST — every claim, finding, and recommendation requires file:line proof or traced evidence with confidence percentage (>80% to act, <80% must verify first).

OOP & DRY Enforcement: MANDATORY IMPORTANT MUST — flag duplicated patterns that should be extracted to a base class, generic, or helper. Classes in the same group or suffix (ex *Entity, *Dto, *Service, etc...) MUST inherit a common base (even if empty now — enables future shared logic and child overrides). Verify project has code linting/analyzer configured for the stack.

External Memory: For complex or lengthy work (research, analysis, scan, review), write intermediate findings and final results to a report file in plans/reports/ — prevents context loss and serves as deliverable.

Quick Summary

Goal: Auto-review a refined PBI artifact for completeness, quality, and correctness before story creation proceeds.

Key distinction: AI self-review (automatic), NOT user interview.

Be skeptical. Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence percentages (Idea should be more than 80%).

Workflow

• Locate PBI — Find latest PBI artifact in team-artifacts/pbis/ or active plan context

• Evaluate checklist — Score each check as PASS/FAIL

• Classify — PASS (all Required + >=50% Recommended), WARN (all Required), FAIL (any Required fails)

• Output verdict — Status, issues, recommendations

Checklist

Required (all must pass)

• Problem statement — Clear problem defined (not just solution description)

• Acceptance criteria — Minimum 3 GIVEN/WHEN/THEN scenarios

• Story points + complexity — Both fields present with valid values

• Dependencies table — Has dependency table with must-before/can-parallel/blocked-by types

• Stakeholder validation — User interview was conducted (validation section present)

• No vague language — No "should work", "might need", "TBD" in acceptance criteria

• Scope boundary — Clear "out of scope" or "not included" section

• Authorization defined — PBI has "Authorization & Access Control" section with roles × CRUD table (ref: .claude/skills/shared/cross-cutting-quality-concerns-protocol.md §1)

Recommended (>=50% should pass)

• RICE/MoSCoW score — Prioritization applied

• Domain vocabulary — Uses project-specific terms from domain-entities-reference.md

• Risk assessment — Risks identified with mitigations

• Non-functional requirements — Performance, security, accessibility considered

• Production readiness concerns — PBI includes "Production Readiness Concerns" table with Yes/No/Existing for: code linting, error handling, loading indicators, Docker integration, CI/CD quality gates (ref: .claude/skills/shared/scaffold-production-readiness-protocol.md )

• Seed data assessed — PBI addresses seed data needs (reference data, config data, test data) or explicitly states "N/A" (ref: .claude/skills/shared/cross-cutting-quality-concerns-protocol.md §2)

• Data migration assessed — PBI addresses schema changes and data migration needs or explicitly states "N/A" (ref: .claude/skills/shared/cross-cutting-quality-concerns-protocol.md §5)

Output

PBI Review Result

Status: PASS | WARN | FAIL Artifact: {pbi-path}

Required ({X}/{Y})

• ✅/❌ Check description

Recommended ({X}/{Y})

• ✅/⚠️ Check description

Issues Found

• ❌ FAIL: {issue}
• ⚠️ WARN: {issue}

Verdict

{PROCEED | REVISE_FIRST}

Key Rules

• FAIL blocks workflow — If FAIL, do NOT proceed to /story. List specific fixes needed.

• WARN allows proceeding — Note gaps but continue.

• No guessing — Every check must reference specific content in the PBI artifact.

• Constructive — Focus on implementation-blocking issues, not pedantic details.

Next Steps

MANDATORY IMPORTANT MUST after completing this skill, use AskUserQuestion to recommend:

• "/story (Recommended)" — Create user stories from validated PBI

• "/refine" — Re-refine if FAIL verdict

• "Skip, continue manually" — user decides

Closing Reminders

MANDATORY IMPORTANT MUST break work into small todo tasks using TaskCreate BEFORE starting. MANDATORY IMPORTANT MUST validate decisions with user via AskUserQuestion — never auto-decide. MANDATORY IMPORTANT MUST add a final review todo task to verify work quality.