Skip to content
Content
Skill

/performance-review-coach

Drafts and critiques performance reviews using Glenn Kramon's seven rules from Stanford GSB's Winning Writing course. Use when the user is writing an annual review, mid-year check-in, self-review, 360 feedback, peer review, or any written feedback for an employee, boss, partner,

From plugin
winning-writing
1731 skills

Claude Code

$ npx -y skills add kalyvask/winning-writing --skill performance-review-coach --agent claude-code

SKILL.md

performance-review-coach.SKILL.md
name: performance-review-coach
description: Drafts and critiques performance reviews using Glenn Kramon's seven rules from Stanford GSB's Winning Writing course. Use when the user is writing an annual review, mid-year check-in, self-review, 360 feedback, peer review, or any written feedback for an employee, boss, partner, or themselves. Triggers on phrases like "performance review," "annual review," "self-review," "write up my feedback for," "review my report's performance," "evaluate my employee," "write feedback for [name]," "promotion packet," "calibration writeup."

Performance-review coach

Source: `points/performance-review-rules.md` and `points/banned-jargon.md`. Read those first.

What this skill does

Given a draft review (or a target person + context), produce a review that the recipient would save if their house were on fire. Most reviews fail by being written ABOUT the person, leading with the negative, ambushing, or psychoanalyzing — this skill catches all four.

The philosophy in one line

A review is a **letter to one person**, not a report card sent up to management. Read aloud, it should sound like you talking to them. If it reads like it's written to their mother, rewrite it.

Mode 1 — Draft from scratch

If the user gives you a person + context, walk through Kramon's seven rules **in order** before writing:

1. **Letter form** — write to them, not about them. *"You ably led..."* not *"Glenn ably led..."* 2. **What you LIKE first** — ask the user for the 2-3 most significant accomplishments. Cite colleagues if any said something specific. Never lead with the negative. Use the ~20:1 positive ratio as a rough check; if every paragraph is critique, rewrite. 3. **What you WOULD LIKE** — convert blunt feedback to *"what I like + what I would like."* See the reframing table in `performance-review-rules.md`. Phrase as a question where possible. 4. **Near-term goals** — specific, named, with what you will look for. 5. **Long-term goals including the next job** — name the role they likely want next; name the 2-3 competencies they need to develop to get there. This is the move that earns the most loyalty. 6. **Colleague feedback** — if you have it, fold it in directly (*"your colleagues describe you as..."*). It lands ten times harder than the boss's own praise. 7. **Vivid examples** — every claim, positive or negative, anchored to a specific moment, project, or behavior. Vague reviews are the worst kind. 8. **Thank you** — close with a specific thank-you. Two words minimum, ideally a paragraph.

Then write. Length target: 400 words for a peer or report; 600+ for a star performer where the long-term-job section is fully built out.

Mode 2 — Critique an existing draft

Mark the draft against the seven rules. Quote the specific lines that fail and rewrite them inline. Flag any of these on sight:

  • **Third-person voice** (*"Peter should strive to..."*) — rewrite to second person.
  • **Lead-with-the-negative** — restructure so the first paragraph is what you like.
  • **Psychoanalysis** (*"your divorce affected your performance," "we understand you've had a tough year"*) — cut entirely. Suggest the user have a private conversation instead.
  • **Ambush** (any negative item the user has not discussed before) — ask the user whether they've raised it. If not, either remove from the review or instruct the user to discuss it with the person *before* finalizing.
  • **The B-word** (*"boring"*) — never. Replace with the specific length issue, specific fix.
  • **Vague verbs** (*"actively involved in," "will continue the evolution"*) — rewrite with named projects, named outcomes.
  • **Passive-aggressive credit** (*"has been actively involved on our team that won four Pulitzers"* — did they win them or not?) — assign credit cleanly.
  • **No long-term horizon** — if the review only covers the current role and the current year, add the next-job paragraph.
  • **Closing with a process note** instead of thank-you — rewrite the ending.

Mode 3 — Self-review

If the user is writing about themselves, ask for:

1. Most significant accomplishments — specific, named, with real numbers. 2. Goals for the current year — what you will deliver, by when. 3. Ambitions for the years beyond, including the job you'd want next.

Then write it in the same letter voice as a peer review, but addressed to the reader's manager. The self-review's job is to give the manager something concrete to react to — not to be modest, not to be a brag sheet.

Mode 4 — Reviewing the weakest performer

The highest-stakes review. The user's whole team is watching this one. Add to the standard mode 1:

  • **Schedule the follow-up explicitly** in the review (*"we'll meet every other Monday to review progress on the goals below"*).
  • **Name the stakes clearly** (*"unless we see improvement on [specific behaviors], we'll need to find you a different role"*) — without this, a later managed-out conversation is unfair.
  • **Behaviors, not character** (*"colleagues say you should stop blaming others"*, not *"you are a blamer"*).
  • **Show you're trying to help** — name what *you* will do: weekly check-ins, mentor intro, scope reduction, training budget.

The angels on your shoulder are the colleagues who want this person to either improve or move on; write the review they would consider fair.

What never goes in a review

  • Speculation about *why* the person had a bad year (divorces, illnesses, mental health, family troubles). Behavior only.
  • Bad news the user has not already raised with the person.
  • Generic adjectives doing the work of evidence (*significant, substantial, robust, comprehensive, impactful*). Cut them or name the number.
  • Sarcasm. Reviewers cannot always tell, and it lives in the file forever.
  • A "next steps" or "look forward to" line as the closing. Close with thank-you.

The reframing kit

For converting blunt manager-speak into review-grade language. (Full table in `performance-review-rules.md`.) Pat

Read more
@kalyvask
Ships withwinning-writing

31 Claude skills for cold outreach, op-eds, pitches, press inquiries, bios, exec memos, performance reviews, spoken-delivery talks, fact-checking, and reply-rate tracking.

Stars on GitHub

Momentum

stars in total
17
Star history of kalyvask/winning-writing on GitHub, a chart by star-history.com

Other skills on winning-writing.

bluf-rewriter

bluf-rewriter

Reorganizes a memo, email, status update, or report so the bottom line is up front. Implements Kramon's BLUF (Bottom Line Up…

compression

compression

Cuts a draft to a target word count without losing substance. Two modes behind one skill. target-count hits a specific number…

fun-angle

fun-angle

Finds the unexpected, dry, or self-deprecating angle that makes an email memorable. Turns "professional and polite" into "this…

More skills from other plugins.

prompt-lookup

prompt-lookup

Activates when the user asks about AI prompts, needs prompt templates, wants to search for prompts, or mentions prompts.chat. Use…

@gh-f@gh-fInstall Skill
ppt-master

ppt-master

AI-driven presentation workflow for generating editable PPTX decks and slides, reconstructing page visuals, creating reusable…

@hugohe3@hugohe3Install Skill
morph-ppt

morph-ppt

Use this skill when the user wants a .pptx with smooth cross-slide animation — PowerPoint Morph transitions, Keynote-style…

officecli12d33k
@iofficeai@iofficeaiInstall Skill

Questions about performance-review-coach.

Which coding agents can use performance-review-coach?
performance-review-coach installs in Claude Code, Codex, Cursor, Gemini CLI, GitHub Copilot, OpenCode and Windsurf with the skills CLI, using each agent's flag shown on this page. Flowy's auto-invocation, which fires a skill when your prompt matches, runs in Claude Code only.