Reviews task-mapping.md and tasks.md for gaps in vertical slice completeness, TDD structure, story coverage, dependencies, and traceability; generates classified issues.
From humaninloopnpx claudepluginhub deepeshbodh/human-in-loop --plugin humaninloopThis skill uses the workspace's default tool permissions.
references/ISSUE-TEMPLATES.mdreferences/PHASE-CHECKLISTS.mdSearches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Searches prompts.chat for AI prompt templates by keyword or category, retrieves by ID with variable handling, and improves prompts via AI. Use for discovering or enhancing prompts.
Executes pre-written implementation plans: critically reviews, follows bite-sized steps exactly, runs verifications, tracks progress with checkpoints, uses git worktrees, stops on blockers.
Find gaps in task artifacts and generate issues that need resolution before proceeding to the next phase. Focus on vertical slice completeness, TDD structure, and traceability. This skill provides phase-specific review criteria for the Devil's Advocate.
humaninloop:validation-plan-artifacts insteadhumaninloop:analysis-specifications insteadEach phase has specific checks to execute. The checks identify Critical, Important, and Minor issues.
| Phase | Focus Area | Key Checks |
|---|---|---|
| Mapping | Story coverage | All P1/P2 stories mapped to cycles |
| Mapping | Slice quality | Cycles are true vertical slices |
| Mapping | Dependencies | Foundation vs feature correctly identified |
| Tasks | TDD structure | Test-first task ordering in each cycle |
| Tasks | Coverage | All cycles have implementation tasks |
| Tasks | Format | Task IDs, file paths, markers correct |
| Cross | Traceability | Stories -> Cycles -> Tasks chain complete |
See PHASE-CHECKLISTS.md for detailed phase-specific checklists and key questions.
Issues are classified by severity to determine appropriate action:
| Severity | Definition | Action |
|---|---|---|
| Critical | Blocks progress; must resolve | Return to Task Architect |
| Important | Significant gap; should resolve | Flag for this iteration |
| Minor | Polish item; can defer | Note for later |
See ISSUE-TEMPLATES.md for severity classification rules, issue documentation formats, and report templates.
Read and understand:
For each check in the phase-specific checklist:
| Verdict | Criteria |
|---|---|
| ready | Zero Critical, zero Important issues |
| needs-revision | 1-3 Important issues, fixable in one iteration |
| critical-gaps | 1+ Critical or 4+ Important issues |
Before finalizing review, verify:
❌ Reviewing task-mapping.md in isolation ✅ Always verify alignment with stories from spec and data model
❌ Marking formatting issues as "Critical" ✅ Reserve Critical for gaps that genuinely block implementation
❌ "The cycles are not vertical slices" ✅ "Cycle 2 groups all database tasks (tasks 2.1-2.4) without user-facing value - horizontal layer, not vertical slice"
❌ Assuming every cycle has proper TDD structure ✅ Explicitly check each cycle starts with a TEST: task before implementation
❌ Accepting tasks without story labels ✅ Every task MUST trace back to a user story via cycle mapping
❌ Reviewing incomplete task artifacts mid-creation ✅ Wait for Task Architect to signal completion before validation