code-review-standards — скилл для Claude Code
Скилл code-review-standards задаёт чек-лист код-ревью с четырьмя уровнями серьёзности и выносит вердикт APPROVE, WARN или BLOCK — чтобы ревью не расползалось на вкусовщину.
Apache-2.0 223 установок ★ 80 обновлён 2026-07-18
npx skills add bobmatnyc/claude-mpm-skills --skill code-review-standards -a claude-codeЧто делает code-review-standards
Скилл задаёт структурированный чек-лист для ревью кода, разбитый по серьёзности. Сначала идёт CRITICAL: секреты в коде, SQL-инъекции, пути исполнения через eval и exec, обходимая авторизация, циклы без выхода. Дальше HIGH — типы у публичных функций, чистый прогон mypy --strict, зелёный pytest, покрытие нового кода не ниже 90%, голые except, мутабельные аргументы по умолчанию, блокирующий ввод-вывод внутри async def.
MEDIUM и LOW идут на заметку: длина функций, таймауты у асинхронных вызовов, SELECT *, мёртвый код, вложенные switch. Для каждого уровня лежит отдельный файл в references (criteria-critical.md, criteria-high.md и остальные), рядом criteria-efficiency.md, criteria-transferability.md и verdict-protocol.md.
Вердикт пишется первой строкой: APPROVE при нуле CRITICAL и HIGH, WARN при HIGH без CRITICAL, BLOCK при любом CRITICAL. Ниже таблица с файлом, строкой, проблемой и требуемой правкой. Работает фильтр уверенности 80%: сомнительное уходит в вопросы к автору, а не в находки, потому что чистый прогон — тоже нормальный результат. Чек-лист при этом заточен под Python (mypy, pytest, PEP 8, asyncio), и в проекте на другом языке часть пунктов HIGH просто повиснет. Вручную скилл не запустить: в нём стоит user-invocable: false, его подгружает агент code-critic.
Какие задачи решает
- поймать хардкод ключа и SQL-инъекцию до мержа
- получить одинаковый вердикт от разных прогонов ревью
- отделить блокирующие находки от замечаний на будущее
- проверить свой код до того, как звать ревьюера
- зафиксировать формат отчёта: вердикт, таблица, правки
Примеры запросов агенту
- Прогони
code-review-standardsпо диффу вauth.pyи выдай вердикт первой строкой. - Проверь новую ветку по чек-листу
code-review-standards, находки ниже 80% уверенности вынеси в вопросы. - Сделай самопроверку по
code-review-standardsперед тем, как зватьcode-critic.
Как установить code-review-standards
1Через Skills CLI (Vercel)
npx skills add bobmatnyc/claude-mpm-skills --skill code-review-standards -a claude-code2Вручную: положить файлы
~/.claude/skills/code-review-standards/SKILL.mdПапка скилла целиком; для одного проекта — .claude/skills/code-review-standards/
1Через Skills CLI (Vercel)
npx skills add bobmatnyc/claude-mpm-skills --skill code-review-standards -a codex2Вручную: положить файлы
.agents/skills/code-review-standards/SKILL.mdДля всех проектов — ~/.agents/skills/
1Через Skills CLI (Vercel)
npx skills add bobmatnyc/claude-mpm-skills --skill code-review-standards -a cursor2Вручную: положить файлы
.cursor/skills/code-review-standards/SKILL.mdCursor читает и .agents/skills/, и .claude/skills/
1GitHub CLI
gh skill install bobmatnyc/claude-mpm-skills universal/process/code-review-standards2Вручную: положить файлы
.github/skills/code-review-standards/SKILL.md1Через Skills CLI (Vercel)
npx skills add bobmatnyc/claude-mpm-skills --skill code-review-standards -a gemini-cli2Командой платформы
gemini skills install https://github.com/bobmatnyc/claude-mpm-skills --path universal/process/code-review-standardsВручную: положить файлы
.devin/skills/code-review-standards/SKILL.mdСтарый путь .windsurf/skills/ тоже читается
Через Skills CLI (Vercel)
npx skills add bobmatnyc/claude-mpm-skills --skill code-review-standards -a opencodeClaude.ai
Упакуйте папку в zip и загрузите в Settings → Capabilities → Skills
По шагам, с командами для каждой платформы: Как установить скилл в Claude Code: пять способов
Вопросы о code-review-standards
Как установить code-review-standards в Claude Code?
Выполните npx skills add bobmatnyc/claude-mpm-skills --skill code-review-standards -a claude-code. Файл скилла окажется в ~/.claude/skills/code-review-standards/SKILL.md.
Что делает code-review-standards?
Даёт агенту ревью фиксированный чек-лист с уровнями CRITICAL, HIGH, MEDIUM и LOW и требует вынести вердикт первой строкой отчёта. CRITICAL и HIGH блокируют поставку, остальное фиксируется как заметки.
Что означают вердикты APPROVE, WARN и BLOCK?
APPROVE — ни одной находки CRITICAL или HIGH. WARN — есть HIGH, но нет CRITICAL, ревью идёт дальше с записанными замечаниями. BLOCK — хотя бы один CRITICAL, пайплайн останавливается.
Исходный текст
SKILL.md · Apache-2.0
---
name: code-review-standards
description: "Severity-tagged code review checklist (CRITICAL/HIGH/MEDIUM/LOW) used by code-critic agent"
user-invocable: false
disable-model-invocation: true
license: Apache-2.0
compatibility: claude-code
progressive_disclosure:
entry_point:
summary: "Severity-tagged checklist for code review. CRITICAL/HIGH block delivery; MEDIUM is flagged; LOW is noted. Critic outputs APPROVE (zero CRITICAL/HIGH), WARN (some HIGH, no CRITICAL), or BLOCK (any CRITICAL). 80% confidence filter — don't manufacture findings."
when_to_use: "Loaded by code-critic agent to apply structured review criteria. Also loaded by engineers to self-check before requesting review."
quick_start: "Apply checklist top-to-bottom: CRITICAL first, then HIGH, MEDIUM, LOW. For each finding: cite file+line, explain why it's a problem, provide the fix. Filter to >80% confidence. Output verdict with finding table."
references:
- criteria-critical.md
- criteria-high.md
- criteria-medium.md
- criteria-low.md
- criteria-efficiency.md
- criteria-transferability.md
- verdict-protocol.md
---
# Code Review Standards
## Purpose
This skill defines the structured checklist that the `code-critic` agent applies during
Stage 4 of the code production pipeline. The checklist is severity-tagged so that PM
and engineer both know exactly which findings block delivery and which are advisory.
Engineers may load this skill for self-review before requesting a critic pass.
The checklist exists because unstructured code review in multi-agent systems produces
inconsistent signal: one critic dispatch flags naming; another flags security; neither
flags the same things. Severity tagging makes the review deterministic across dispatches.
## The Severity-Tagged Checklist
### CRITICAL (must fix, blocks delivery)
- [ ] No secrets, API keys, or credentials hardcoded
- [ ] No SQL injection vectors (parameterized queries only)
- [ ] No arbitrary code execution paths (no `eval`, `exec`, unrestricted `pickle.loads`)
- [ ] Authentication/authorization not bypassable
- [ ] No infinite loops without escape conditions
### HIGH (must fix, blocks delivery)
- [ ] Type hints on all public functions and classes
- [ ] mypy --strict passes with zero errors
- [ ] pytest passes with zero failures
- [ ] Test coverage >= 90% on new code
- [ ] No bare except clauses
- [ ] No mutable default arguments
- [ ] No global mutable state
- [ ] No synchronous I/O inside async functions
- [ ] No N+1 query patterns
- [ ] Error cases handled explicitly (not silently swallowed)
### MEDIUM (flag, note in report, proceed)
- [ ] Functions <= 20 lines (prefer <= 10)
- [ ] No nested loops where hash map would reduce complexity
- [ ] list.pop(0) replaced with deque.popleft() where relevant
- [ ] asyncio.gather uses return_exceptions=True where appropriate
- [ ] Async operations have explicit timeouts
- [ ] Docstrings on public methods (Google or NumPy style)
- [ ] No Any types in production code paths
**Efficiency** (see [criteria-efficiency.md](references/criteria-efficiency.md)):
- [ ] No nested loops over two collections that should be a hash-map lookup (O(n*m) → O(n+m))
- [ ] No per-iteration I/O (queries/RPCs/fetches inside a loop) — batch outside the loop (HIGH on hot paths; see "No N+1 query patterns")
- [ ] Repeated deep property/selector resolution cached in a local (no greedy data access)
- [ ] String accumulation in loops uses list+join / StringBuilder, not `+=`
- [ ] No `SELECT *` / over-fetching in production query paths
**Transferability** (see [criteria-transferability.md](references/criteria-transferability.md)):
- [ ] No dead/unreachable code (statements after unconditional return/break/raise; uncalled private members)
- [ ] Long `if/else if` chains (3+ branches on one key) replaced by switch/match or dispatch map
- [ ] No nested `switch`/`match` — extract inner switch to a named function
- [ ] No actively misleading names (name contradicts the value it holds)
### LOW (note only)
- [ ] PEP 8 compliance (black + isort handles this automatically)
- [ ] Variable naming is clear and descriptive
- [ ] No commented-out code left in
- [ ] Import ordering is clean
- [ ] Naming consistency across the change (same concept, same name; consistent casing)
- [ ] No file managing too many responsibilities (see 800-line file limit)
## Verdict Format
Critic output MUST begin with the verdict on the first line, followed by the finding
table, followed by a summary paragraph.
**First line format:**
```
VERDICT: APPROVE
```
or
```
VERDICT: WARN
```
or
```
VERDICT: BLOCK
```
**Finding table format:**
| Severity | File | Line | Issue | Required Fix |
|----------|------|------|-------|--------------|
| CRITICAL | auth.py | 47 | Hardcoded API key `sk-...` | Move to env var; add to `.env.example` |
| HIGH | fetcher.py | 23 | `requests.get()` called inside `async def` | Replace with `await httpx.AsyncClient().get()` |
| MEDIUM | parser.py | 88 | Function is 34 lines | Extract `_parse_headers()` helper |
**Verdict definitions:**
| Verdict | Condition | PM Action |
|---------|-----------|-----------|
| APPROVE | Zero CRITICAL, zero HIGH findings | Proceed to Stage 5 (Security) |
| WARN | Zero CRITICAL, one or more HIGH findings | Proceed to Stage 5 with findings logged to docs handoff |
| BLOCK | Any CRITICAL finding (one or more) | Halt pipeline; surface findings to user; await user direction |
**APPROVE** means the implementation is ready for security review. MEDIUM and LOW findings
in an APPROVE review are passed to the Documentation agent as notes — they do not block
delivery but are preserved for future reference.
**WARN** means the implementation has structural issues that should be fixed but do not
represent exploitable defects or correctness failures. PM proceeds to security review
and appends the WARN finding table to the documentation handoff message. PM also logs
the findings (KB entry or todo) so they are not silently dropped.
**BLOCK** means the implementation has at least one defect that, if shipped, creates a
security vulnerability, data loss risk, or silent failure mode. PM halts the pipeline
immediately, presents the critic finding table verbatim to the user, and awaits explicit
direction. PM MUST NOT auto-retry the engineer without user input.
## 80% Confidence Filter
A clean review is a valid review. Do not manufacture findings.
Only report issues with >80% confidence they are real problems. Do not flag:
- Style preferences as HIGH or CRITICAL
- "This could theoretically be an issue in an edge case" without specific evidence
- Patterns that look unusual but may be intentional and correct
- Missing features that were not in scope (check the Stage 1 spec)
When confidence is below 80%, note the concern as a question in the summary paragraph
rather than as a finding in the table. Example: "The `process_batch()` function did not
appear to handle empty input — verify whether the caller guarantees non-empty batches."
This filter prevents the critic from becoming a noise generator that trains PM to
ignore findings. Each finding in the table should be actionable: engineer reads it,
knows exactly what to fix, and can do so without asking for clarification.
## Navigation
For detailed criteria with examples:
- **[CRITICAL Criteria](references/criteria-critical.md)**: Detailed explanations and examples for each CRITICAL item
- **[HIGH Criteria](references/criteria-high.md)**: Detailed explanations and examples for each HIGH item
- **[MEDIUM Criteria](references/criteria-medium.md)**: Detailed explanations and examples for each MEDIUM item
- **[LOW Criteria](references/criteria-low.md)**: Detailed explanations and examples for each LOW item
- **[Efficiency Criteria](references/criteria-efficiency.md)**: Algorithmic/data-access patterns (nested loops, fetch-in-loop, greedy access, loop concatenation, over-fetching)
- **[Transferability Criteria](references/criteria-transferability.md)**: Maintainability patterns (dead code, long if/else-if chains, nested switch, naming hygiene)
- **[Verdict Protocol](references/verdict-protocol.md)**: Full PM behavior for each verdict, failure loop templates
> The Efficiency and Transferability criteria are derived from CAST Highlight code
> quality indicators (https://doc.casthighlight.com/), paraphrased with original
> examples.
© bobmatnyc · Apache-2.0 · bobmatnyc/claude-mpm-skills