Agent skill
review-deferred
Review deferred verification items. Use when you want to check or clear the deferred review queue without waiting for project completion.
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/review-deferred
SKILL.md
Review the deferred verification queue and optionally clear reviewed items.
Context Detection
Determine working context:
-
If current working directory matches pattern
*/features/*:- PROJECT_ROOT = parent of parent of CWD (e.g.,
/project/features/foo→/project) - MODE = "feature"
- PROJECT_ROOT = parent of parent of CWD (e.g.,
-
Otherwise:
- PROJECT_ROOT = current working directory
- MODE = "greenfield"
Workflow
Copy this checklist and track progress:
Review Deferred Progress:
- [ ] Step 1: Read deferred queue
- [ ] Step 2: Present items grouped by phase
- [ ] Step 3: Collect user decisions
- [ ] Step 4: Update queue file
Step 1: Read Deferred Queue
Read .claude/deferred-reviews.json from PROJECT_ROOT.
If file doesn't exist or queue is empty:
NO DEFERRED ITEMS
=================
The deferred review queue is empty. No items pending human review.
Stop here if no items to review.
If file fails to parse:
Back up the corrupt file to .claude/deferred-reviews.json.bak, report the issue,
and stop:
QUEUE FILE CORRUPT
==================
.claude/deferred-reviews.json could not be parsed.
Backed up to .claude/deferred-reviews.json.bak
Run /go to continue — a fresh queue will be created automatically.
Step 2: Present Items
Filter to items where reviewed: false. Group by phase:
DEFERRED REVIEW QUEUE
=====================
{N} items pending review across {P} phases.
Phase 1:
- [ ] "{criterion}" (Task {task_id})
Context: {file}:{line} — {summary}
Deferred: {deferred_at}
Phase 3:
- [ ] "{criterion}" (Task {task_id})
Context: {file}:{line} — {summary}
Deferred: {deferred_at}
Step 3: Collect User Decisions
Use AskUserQuestion with options:
- "All look good" — Mark all items as
reviewed: true, then clear - "Review individually" — Present each item one by one (see Individual Review below)
- "Cancel" — No changes, keep queue as-is
Individual Review
For each unreviewed item, use AskUserQuestion:
Item {i}/{N}: "{criterion}" (Task {task_id})
Context: {file}:{line} — {summary}
Options:
- "Looks good" — Mark
reviewed: true - "Needs fix" — Keep
reviewed: false, note for follow-up - "Skip" — Leave unchanged, move to next item
After reviewing all items, summarize:
REVIEW COMPLETE
===============
Approved: {X}
Needs fix: {Y}
Skipped: {Z}
Step 4: Update Queue File
- Set
reviewed: trueon approved items - Remove all
reviewed: trueitems from the queue array - Update
last_drainedandlast_drain_reasonfields:json{ "last_drained": "{ISO timestamp}", "last_drain_reason": "explicit_review" } - Write using atomic pattern: write to temp file, rename to
.claude/deferred-reviews.json
If items marked "Needs fix" exist:
FOLLOW-UP NEEDED
================
{Y} items need attention:
- "{criterion}" (Task {task_id})
Context: {file}:{line}
These items remain in the queue and will surface again at the next drain trigger.
Drain Reasons
When invoked by other skills, the drain reason is passed as context:
| Caller | Drain Reason |
|---|---|
/review-deferred (direct) |
explicit_review |
/go Case E (completion) |
project_completion |
/go Case F (blocker) |
blocker_drain |
/phase-checkpoint (max queue) |
max_queue_reached |
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
Didn't find tool you were looking for?