Menu
Processes review feedback until the change is approved.
Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Basé sur la classification professionnelle SOC
Flags risky shell commands and unsafe tree ops.
Detects high-confidence security risks in code.
Surfaces the dojo's non-negotiable prime directives.
Activates the dojo framework at the start of a session.
Plans multi-step work before writing code.
Captures lessons and promotes recurring patterns.
| name | receiving-code-review |
| description | Processes review feedback until the change is approved. |
| tier | practical |
| category | workflow |
| created_by | human |
| platforms | ["windows","macos","linux"] |
| tags | ["review","feedback","iteration"] |
| author | Andreas Wasita (@andreaswasita) |
Processes every review comment — human, agent, or CI — until the change is approved. Each comment is categorized, addressed, verified, and answered. Does NOT silently ignore comments and does NOT amend the original commit during active review.
requesting-code-review skill flagged issues.view, edit, grep, and powershell Copilot tools.git for new commits and gh for PR threads (when relevant).1. Read every comment. Categorize by action.
2. For each comment: fix, verify, respond, mark resolved.
3. Re-run the full test suite after all fixes.
4. Re-self-review with the `requesting-code-review` skill.
5. Update `tasks/todo.md` and (if pattern recurring) `tasks/lessons.md`.
6. Request re-review with a short summary.
| Category | Action |
|---|---|
| Agree — will fix | Fix it. No debate. |
| Agree — won't fix now | Log as follow-up with reason |
| Disagree — with evidence | Respond with reasoning and evidence |
| Need clarification | Ask one focused follow-up |
| Already addressed | Link to the commit / line |
| Phase | Tool |
|---|---|
| Read comments | gh pr view / PR UI |
| Apply fix | edit |
| Verify fix | powershell → test runner |
| Respond | gh pr comment / PR UI |
| Track resolution | Checklist in tasks/todo.md |
Read every comment. Categorize each using the table above. Track them in tasks/todo.md as a Review Feedback section:
### Review Feedback (Round 1)
- [ ] Comment #1: missing error handling — agree, will fix
- [ ] Comment #2: rename for clarity — agree, will fix
- [ ] Comment #3: add pagination — agree, follow-up (out of scope)
For every comment:
edit.powershell tool.tasks/todo.md.### Comment #1: missing error handling — @reviewer
**Action:** Added try/catch with 502 response on upstream failure (commit `abc1234`).
**Verification:** New test covers the failure path.
Run the full test suite — not just the tests for the changed lines. Re-run the dojo gate:
bash scripts/verify.sh --check
Re-self-review with the requesting-code-review skill. Every fix is a potential new bug.
If review keeps catching the same shape of issue, append to tasks/lessons.md:
- date: 2026-05-19
error_type: missing-error-handling
trigger: "Reviewer caught missing try/catch on external API call"
root_cause: "Did not consider failure modes for external dependencies"
fix: "Wrapped call in retry + circuit breaker"
rule: "Every external API call has an explicit failure path with a test"
occurrences: 2
status: active
**Re-review ready**
- Addressed 2/3 comments directly
- 1 logged as follow-up (out of scope)
- All tests pass (49 total, 2 new)
- Self-review clean
Push fixes as new commits while review is active. Squash fixup commits just before merge — not during.
tasks/todo.md.requesting-code-review re-run after fixes; no new 🔴 findings.tasks/lessons.md was updated.