Claude Prompt Library

Claude Prompts for Performance Reviews That Aren't Generic

20 copy-paste prompts

20 copy-paste Claude prompts for writing performance reviews, synthesizing 360 feedback, calibrating ratings, drafting self-reviews, and the difficult conversations the review surfaces. Long-context Claude shines when fed full context.

Writing Reviews

4 prompts

Performance Review from Notes

1/20

[Paste notes from year — 1:1s, projects, peer feedback]. Write a performance review draft for [employee]. Include: summary paragraph, strengths with specific examples (2-3), growth areas with specific examples (2-3), goals achievement vs commitment, calibration notes for HR. Specific over generic. Use my voice from notes — not corporate-speak.

Drafts performance reviews from your notes.

💡

Pro tip: Generic reviews ("strong performer, continues to grow") add zero value + waste reader's time. Specific examples = review reader trusts the rating + knows what to do with it.

Strength Articulation

2/20

Help me articulate [employee]'s strength. I'll describe what they do well. Output: name the strength precisely (not "great team player"), describe the behavior with examples, why it matters for the role, where it could compound if developed. Make it concrete enough that another manager would recognize it.

Articulates strengths concretely.

💡

Pro tip: "Great team player" = useless. "Brings hidden disagreements into 1:1s before team meetings, so debates happen on substance not surprise" = concrete + transferable. Specificity = dignity.

Growth Area Without Crushing

3/20

Help me write a growth area for [employee]. The behavior: [describe]. Why it matters: [describe]. Output the feedback: name behavior specifically, impact (concrete examples), what better looks like, what support I'll give, my belief in their ability to grow. Honest but not harsh.

Writes growth feedback that motivates.

💡

Pro tip: Growth feedback without "what better looks like" = paralysis. Without "support I'll give" = abandonment. Without "I believe you can" = career-threat read. All three = development.

Goals vs Commitments Review

4/20

[Employee] committed to [goals at start of year]. Outcomes: [describe]. Write the goals section: per goal, what was committed, what was delivered, gap or over-delivery, what we learned about goal-setting + scope. Don't penalize stretch goals not fully hit if reach was real.

Reviews goals vs commitments fairly.

💡

Pro tip: Conservative goals fully met < stretch goals 70% met. Penalizing stretch attempts = next year's sandbagged goals. Reward reach + clarify what we learned.

XML tags are just the start. Learn the full Claude workflow.

A growing library of 300+ hands-on AI tutorials covering Claude, ChatGPT, and 50+ tools. New tutorials added every week.

Start 7-Day Free Trial

360 + Calibration

3 prompts

360 Feedback Synthesis

5/20

[Paste 360 feedback — peer, report, cross-functional]. Synthesize for [employee]'s review. Output: themes (2-3 strengths, 2-3 growth areas), where feedback converges (high-confidence signal), where it diverges (context-dependent or rater-bias-dependent), what's actionable. Don't average — interpret.

Synthesizes 360 feedback for reviews.

💡

Pro tip: Averaging 360 feedback = lose signal. "3 peers said X, 1 said opposite" = X is signal + opposite is context (different rater experience? rater bias?). Interpret, don't average.

Calibration Prep

6/20

I'm going into calibration with [team]. My ratings: [list]. Help me prep: rationale for each rating (1-2 sentences each), where I might face pushback, who I want to advocate harder for, where I'm willing to compromise, where I'm not (and why). Anticipate other managers' arguments.

Preps for performance calibration.

💡

Pro tip: Calibration is political AND legitimate. Going in with rationale + anticipated pushback = effective advocacy. Going in unprepared = your team rated by who advocates loudest.

Disagreement with HR Rating

7/20

HR is pushing back on my [rating] for [employee]. HR's reason: [describe]. Help me respond: their concern reframed in best light, my evidence + rationale, where I might be wrong (steel-man their case), where I'm clearly right (hold the line), compromise option if any. Professional disagreement.

Disagrees professionally with HR ratings.

💡

Pro tip: HR often has data managers don't (legal exposure, calibration across org, comp implications). Steel-man their case first = professional + sometimes you discover you're wrong. Reflexive defense = career capital wasted.

Self-Reviews + Conversations

3 prompts

Self-Review Draft

8/20

Help me write my self-review. Year accomplishments: [list]. Challenges: [list]. Output: structured by review form headers, self-honest about strengths, self-honest about gaps (without false modesty), specific examples, alignment with role + level expectations, my goals for next year. Confident not arrogant; honest not self-deprecating.

Drafts honest self-reviews.

💡

Pro tip: Self-review tone tightrope: confident enough to advocate for yourself, humble enough to be credible. False modesty + bragging both undermine. Specific examples = automatic right tone.

Self-Review Disagreement Response

9/20

My manager's review of me said [paste section]. I disagree with [specific point]. Help me respond professionally: acknowledge their perspective, present my counter-evidence specifically, ask question rather than declare ("can you share examples of X?"), propose path forward. Disagree without burning bridges.

Responds to review disagreements.

💡

Pro tip: Disagreement-with-rating happens. Done well = career-strengthening (manager updates view). Done poorly = career-limiting (manager sees you as defensive). Specific counter-evidence + open questions = the difference.

Promotion Case

10/20

I'm up for promotion to [next level]. Help me build the case. Output: [next level] expectations vs my current behavior with examples, scope of impact (broader than role), peer ratings or feedback supporting, gaps to acknowledge (don't pretend they don't exist), specific commitment if promoted. 1 page.

Builds promotion cases.

💡

Pro tip: Promotion cases hide gaps = HR + leveling sees through. Acknowledge gap + plan = credibility. Pretending = sandbagged promo. Best cases address weakness directly.

These prompts give you the what. Tutorials give you the why.

Learn when to use extended thinking, how to build Claude Projects, and workflows that compound. 300+ tutorials and growing.

Try AI Academy Free

Difficult Reviews + Edge Cases

4 prompts

Underperformance Review

11/20

[Employee] is underperforming. Help me write the review honestly. Output: clear statement of where performance falls short, specific examples (not vague feelings), expected vs actual, support I've given, what improvement looks like, timeline (formal PIP if needed), my belief about whether they can succeed in role. Honest path — saving face hurts both of us.

Writes underperformance reviews.

💡

Pro tip: Diplomatic underperformance reviews = employee surprised by termination 6 months later. Honest reviews = employee can either rise to it OR self-select out. Either outcome better than dragged-out denial.

Strong Performer Stuck Review

12/20

[Employee] is a strong performer but has plateaued. Help me write the review: acknowledge sustained excellence, name the plateau without being mean, what the next ceiling looks like + why they haven't reached it (capability? interest? scope?), the choice ahead (push for next level vs settle here long-term).

Reviews plateaued strong performers.

💡

Pro tip: Plateau ≠ failure. But unaddressed plateau = quiet quit or surprise resignation. Naming it openly + offering choice (climb OR coast w/ different expectations) = respect + retention.

New Hire (<1 Year) Review

13/20

[Employee] joined [N months ago]. Help me write a partial-year review. Output: ramp progress vs expectations at this tenure, what they're doing well, what to watch for next 6 months, projects ahead, how I'll support, mutual fit assessment. Don't rate as if they had full year — different baseline.

Writes new-hire performance reviews.

💡

Pro tip: Comparing 6-month employee to 6-year employee = unfair to both. Adjust expectation by tenure. "Ramping on track" or "ahead of curve" or "struggle areas to watch" — distinct from full-year ratings.

Manager-of-Manager Review

14/20

[Manager I oversee] is up for review. Help me assess: their team's outcomes (lagging indicator), their team health (1:1 satisfaction, retention, growth), upward feedback signal, peer feedback, executive presence, leadership of leaders skills. Different from IC review — measured by team's output, not their own.

Reviews managers of managers.

💡

Pro tip: Manager performance ≠ manager output. Their team's output IS their output. Strong manager with weak team = haven't leveled up team. Weak manager with strong team = inherited well, hasn't hurt yet. Different lens.

Frequently Asked Questions

Yes for drafting, brainstorming, synthesis. Not for replacing your judgment. Manager reads + edits + owns the review. Your assessment, AI-helped writing. Submitting unread AI output = ethical fail + likely caught for being generic.
Yes if you give Claude no context. Specific notes from the year + your voice = personalized. Vague prompts get vague output. Garbage in, garbage out applies hardest to people writing.
Different views. Some orgs require disclosure; some don't. The deeper question: did the review reflect your judgment? If yes, AI was a writing tool (like spell-check). If no, you owe a real review. Disclosure follows the actual ownership.
Standard consumer Claude doesn't train on conversations (per current policy — verify). For employee data privacy: use Anthropic via Bedrock with controls, or de-identify (initials, role only) before pasting. Org policy may restrict regardless.
HR systems gate-keep the form. They don't help write better content. AI helps the writing; the system handles the storing. Different jobs.

Prompts are the starting line. Tutorials are the finish.

A growing library of 300+ hands-on tutorials on ChatGPT, Claude, Midjourney, and 50+ AI tools. New tutorials added every week.

7-day free trial. Cancel anytime.