performance-review — скилл для Claude Code: что делает
Формирует шаблоны для самооценки, ревью сотрудников и калибровки оценок команды.
npx skills add anthropics/knowledge-work-plugins --skill performance-review -a claude-codeЧто делает performance-review
Скилл генерирует структурированные шаблоны для оценки работы сотрудников. Он поддерживает три режима: self-assessment для личного отчёта, manager для ревью подчинённого и calibration для подготовки к встрече по распределению грейдов. Если запустить команду без аргументов, агент уточнит, какой именно формат требуется.
В шаблоны заложены таблицы целей, списки достижений с фиксацией влияния на результат и матрицы развития навыков. Если подключены коннекторы к HRIS или трекеру задач, инструмент подтягивает историю целей и закрытые тикеты. Без коннекторов факты придётся вводить вручную.
Какие задачи решает
- составить структурированную самооценку за отчётный период
- написать ревью на подчинённого с планом развития
- подготовить распределение оценок и кейсы на повышение к калибровке
- перевести общие впечатления в конкретные поведенческие примеры
Примеры запросов агенту
- /performance-review self-assessment
- /performance-review manager Alex
- /performance-review calibration
Как установить performance-review
1Через Skills CLI (Vercel)
npx skills add anthropics/knowledge-work-plugins --skill performance-review -a claude-code2Вручную: положить файлы
~/.claude/skills/performance-review/SKILL.mdПапка скилла целиком; для одного проекта — .claude/skills/performance-review/
1Через Skills CLI (Vercel)
npx skills add anthropics/knowledge-work-plugins --skill performance-review -a codex2Вручную: положить файлы
.agents/skills/performance-review/SKILL.mdДля всех проектов — ~/.agents/skills/
1Через Skills CLI (Vercel)
npx skills add anthropics/knowledge-work-plugins --skill performance-review -a cursor2Вручную: положить файлы
.cursor/skills/performance-review/SKILL.mdCursor читает и .agents/skills/, и .claude/skills/
1GitHub CLI
gh skill install anthropics/knowledge-work-plugins performance-review2Вручную: положить файлы
.github/skills/performance-review/SKILL.md1Через Skills CLI (Vercel)
npx skills add anthropics/knowledge-work-plugins --skill performance-review -a gemini-cli2Командой платформы
gemini skills install https://github.com/anthropics/knowledge-work-plugins --path human-resources/skills/performance-reviewВручную: положить файлы
.devin/skills/performance-review/SKILL.mdСтарый путь .windsurf/skills/ тоже читается
Через Skills CLI (Vercel)
npx skills add anthropics/knowledge-work-plugins --skill performance-review -a opencodeClaude.ai
Упакуйте папку в zip и загрузите в Settings → Capabilities → Skills
Из того же набора133
Весь набор →Похожие
Вопросы о performance-review
Как установить performance-review в Claude Code?
Выполните команду npx skills add anthropics/knowledge-work-plugins --skill performance-review -a claude-code. Также можно вручную сохранить файл в ~/.claude/skills/performance-review/SKILL.md.
Что делает скилл performance-review?
Он создаёт бланки самооценки, формы оценки подчинённых менеджером и документы для калибровки оценок команды с планами развития.
Как скилл использует внешние интеграции?
При наличии коннектора к HRIS он извлекает историю прошлых ревью и текущие роли. Подключение трекера задач позволяет привязать достижения к закрытым тикетам и вехам проектов.
Исходный текст
SKILL.md · Apache-2.0
--- name: performance-review description: Structure a performance review with self-assessment, manager template, and calibration prep. Use when review season kicks off and you need a self-assessment template, writing a manager review for a direct report, prepping rating distributions and promotion cases for calibration, or turning vague feedback into specific behavioral examples. argument-hint: "<employee name or review cycle>" --- # /performance-review > If you see unfamiliar placeholders or need to check which tools are connected, see [CONNECTORS.md](../../CONNECTORS.md). Generate performance review templates and help structure feedback. ## Usage ``` /performance-review $ARGUMENTS ``` ## Modes ``` /performance-review self-assessment # Generate self-assessment template /performance-review manager [employee] # Manager review template for a specific person /performance-review calibration # Calibration prep document ``` If no mode is specified, ask what type of review they need. ## Output — Self-Assessment Template ```markdown ## Self-Assessment: [Review Period] ### Key Accomplishments [List your top 3-5 accomplishments this period. For each, describe the situation, your contribution, and the impact.] 1. **[Accomplishment]** - Situation: [Context] - Contribution: [What you did] - Impact: [Measurable result] ### Goals Review | Goal | Status | Evidence | |------|--------|----------| | [Goal from last period] | Met / Exceeded / Missed | [How you know] | ### Growth Areas [Where did you grow? New skills, expanded scope, leadership moments.] ### Challenges [What was hard? What would you do differently?] ### Goals for Next Period 1. [Goal — specific and measurable] 2. [Goal] 3. [Goal] ### Feedback for Manager [How can your manager better support you?] ``` ## Output — Manager Review ```markdown ## Performance Review: [Employee Name] **Period:** [Date range] | **Manager:** [Your name] ### Overall Rating: [Exceeds / Meets / Below Expectations] ### Performance Summary [2-3 sentence overall assessment] ### Key Strengths - [Strength with specific example] - [Strength with specific example] ### Areas for Development - [Area with specific, actionable guidance] - [Area with specific, actionable guidance] ### Goal Achievement | Goal | Rating | Comments | |------|--------|----------| | [Goal] | [Rating] | [Specific observations] | ### Impact and Contributions [Describe their biggest contributions and impact on the team/org] ### Development Plan | Skill | Current | Target | Actions | |-------|---------|--------|---------| | [Skill] | [Level] | [Level] | [How to get there] | ### Compensation Recommendation [Promotion / Equity refresh / Adjustment / No change — with justification] ``` ## Output — Calibration ```markdown ## Calibration Prep: [Review Cycle] **Manager:** [Your name] | **Team:** [Team] | **Period:** [Date range] ### Team Overview | Employee | Role | Level | Tenure | Proposed Rating | Notes | |----------|------|-------|--------|-----------------|-------| | [Name] | [Role] | [Level] | [X years] | [Rating] | [Key context] | ### Rating Distribution | Rating | Count | % of Team | Company Target | |--------|-------|-----------|----------------| | Exceeds Expectations | [X] | [X]% | ~15-20% | | Meets Expectations | [X] | [X]% | ~60-70% | | Below Expectations | [X] | [X]% | ~10-15% | ### Calibration Discussion Points 1. **[Employee]** — [Why this rating may need discussion, e.g., borderline, first review at level, recent role change] 2. **[Employee]** — [Discussion point] ### Promotion Candidates | Employee | Current Level | Proposed Level | Justification | |----------|-------------|----------------|---------------| | [Name] | [Current] | [Proposed] | [Evidence of next-level performance] | ### Compensation Actions | Employee | Action | Justification | |----------|--------|---------------| | [Name] | [Promotion / Equity refresh / Market adjustment / Retention] | [Why] | ### Manager Notes [Context the calibration group should know — team changes, org shifts, project impacts] ``` ## If Connectors Available If **~~HRIS** is connected: - Pull prior review history and goal tracking data - Pre-populate employee details and current role information If **~~project tracker** is connected: - Pull completed work and contributions for the review period - Reference specific tickets and project milestones as evidence ## Tips 1. **Be specific** — "Great job" isn't feedback. "You reduced deploy time 40% by implementing the new CI pipeline" is. 2. **Balance positive and constructive** — Both are essential. Neither should be a surprise. 3. **Focus on behaviors, not personality** — "Your documentation has been incomplete" vs. "You're careless." 4. **Make development actionable** — "Improve communication" is vague. "Present at the next team all-hands" is actionable.
© anthropics · Apache-2.0 · anthropics/knowledge-work-plugins