⚡AgentSkills
🧭 Engineering Leadership · People Management

Calibrate performance reviews fairly

Evidence-based reviews with bias checks and consistent leveling across teams.

advanced~45 minEngineering ManagersTech LeadsStaff Engineers

Steps

  1. 1Collect evidence continuously, not from last-month memory
  2. 2Score against written level expectations, not peer comparison
  3. 3Check for bias patterns: recency, likability, visibility of work
  4. 4Calibrate across managers with anonymized examples
  5. 5Separate compensation conversations from growth conversations
  6. 6No surprises rule: anything in the review was said when observed

Common Pitfalls

  • ▲Rockstar-effect inflating visible work over critical invisible work
  • ▲Reviews as annual surprise attack

Commands

Install with skills CLI
$ npx skills add aniruddhaadak80/skills --skill people-management-performance-calibration
Install globally
$ npx skills add aniruddhaadak80/skills --skill people-management-performance-calibration -g

Tags

#performance#reviews#fairness#engineering-leadership#people-management

Related skills

Run one-on-ones that build trustfoundation

Weekly direct-report meetings owned by them: growth, blockers, feedback — not status updates.

🧭 Engineering Leadership·~25m
Open →