Verifies code implementation against plan document: reads plan from disk, inspects codebase independently, runs full tests. Outputs structured PASS/FAIL review to file. Use post-execution.
npx claudepluginhub tmdgusya/engineering-discipline --plugin engineering-disciplineThis skill uses the workspace's default tool permissions.
Independently verifies implementation results using only the plan document and the codebase. Receives no information from the execution process.
Validate that an implementation plan was correctly executed by verifying success criteria and identifying deviations. Use after implementing a plan to verify correctness.
Audits post-implementation plans against changes by parsing checkboxes for completeness and accuracy. Supports autonomy modes (Autopilot, Critical, Verbose) and file-review. Invoke via /verify-plan.
Verifies batch code implementation exactly matches plan specifications for adherence, not quality. Use after plan batch execution, during /cipherpowers:verify execute, or before next batch.
Share bugs, ideas, or general feedback.
Independently verifies implementation results using only the plan document and the codebase. Receives no information from the execution process.
The reviewer shares no memory with the executor. The plan's stated goals and the current state of the codebase — these two alone are the basis for judgment.
plan-crafting first)The only input to this skill is the plan file path.
docs/engineering-discipline/plans/YYYY-MM-DD-<feature-name>.md
The following must never be provided as input:
Use the extracted results as the foundation for the review document.
Inspect the codebase against the files specified in the plan.
git logCombine results from Phases 2–4 to reach a verdict.
PASS conditions (all must be met):
FAIL (if any of the following apply):
After reaching a verdict, write and save the review document.
docs/engineering-discipline/reviews/YYYY-MM-DD-<feature-name>-review.md
(User preferences for review location override this default.)
# [Feature Name] Review
**Date:** YYYY-MM-DD HH:MM
**Plan Document:** `docs/engineering-discipline/plans/YYYY-MM-DD-<feature-name>.md`
**Verdict:** PASS / FAIL
---
## 1. File Inspection Against Plan
| Planned File | Status | Notes |
|---|---|---|
| `path/to/file` | OK / Missing / Mismatch | Details |
## 2. Test Results
| Test Command | Result | Notes |
|---|---|---|
| `pytest tests/...` | PASS / FAIL | Error details if failed |
**Full Test Suite:** PASS / FAIL (N passed, M failed)
## 3. Code Quality
- [ ] No placeholders
- [ ] No debug code
- [ ] No commented-out code blocks
- [ ] No changes outside plan scope
**Findings:**
- (Describe with file path and line number)
## 4. Git History
| Planned Commit | Actual Commit | Match |
|---|---|---|
| `feat: add X` | `abc1234 feat: add X` | OK / Mismatch |
## 5. Overall Assessment
(Summary of the overall judgment. If FAIL, describe specifically which items failed and why.)
## 6. Follow-up Actions
- (If FAIL: list of items that need to be fixed)
- (If PASS: record improvement suggestions if any)
Stop immediately and notify the user in the following situations:
When in doubt, do not guess — ask the user.
| Anti-Pattern | Why It Fails |
|---|---|
| Reading run-plan execution logs to verify | Information isolation violation. Anchors on the executor's framing |
| Trusting previous test results instead of running tests | Environment may have changed after execution. Not independent verification |
| Finding issues and fixing them directly | Violates separation of reviewer and implementer roles |
| Giving a "close enough, PASS" verdict | No conditional passes. If criteria are not met, it is FAIL |
| Delivering review results verbally without saving a document | No verification record remains. Untraceable |
| Judging by criteria not in the plan | The reviewer judges only by the plan's criteria. Adding arbitrary standards is prohibited |
| Receiving a plan summary and verifying from that | Information is lost during summarization. The original must be read directly |
Self-check when review is complete:
After review is complete:
run-plan or systematic-debugging skillplan-crafting skill to revise the planThis skill itself does not invoke the next skill. It saves the review document, reports results, and lets the user decide the next step.