Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when a leader needs to evaluate a marketing team member using data: a KPI scorecard by role, strengths with real examples, areas to improve stated as observable behavior, a 30/90-day development plan, and a 1-5 rating rubric. Trigger on 'performance review', 'evaluate a team member', 'review my marketing team', 'monthly 1:1 review', 'feedback for a team member', or 'development plan for an employee'.
.claude/skills/minhnv0807-65-team-performance-review-global/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 52% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 8% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 22% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 157% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 9% | 0% |
> A review is not a grade — it exists so the person knows what to do to get better. Feedback without data is opinion; feedback with data is coaching. There should be no surprises in a review: feedback is given continuously through the period, and the review only summarizes it.
Read the brief history (64-team-brief-global), last period's output, and the reports (07-marketing-report-global) if available. If missing, ask up to 4 questions:
In many jurisdictions, written performance documentation carries employment-law weight and can be produced as evidence in a dispute. Write it accordingly:
| Role | Quantitative KPIs | Qualitative KPIs | |------|-------------------|------------------| | Content writer | Pieces per month, on-time rate, revisions per piece, engagement rate on their work | On-brand voice, sharpness of the angle, proactively proposing angles | | Designer | Assets per month, on-time rate, share approved on the first pass, design review score | Understands briefs, follows the brand guideline, handles asset gaps | | Media buyer | CPA vs target, ROAS, share of campaigns hitting KPI, optimization frequency | Tracking discipline, quality of the test log, flags risks early | | Channel owner | Response SLA, share of channels set up to checklist, open and click rates | Incident handling, cross-functional coordination |
If no KPI was set for the role at the start of the period, the first action after the review is to set KPIs for the next period. Do not measure someone against a standard they never saw.
markdown# Performance Review — [Name] — [Period] Role: [role] · Reviewer: [leader] · Date: [date] ## I. KPI scorecard | KPI set at period start | Target | Actual | % achieved | Note | |-------------------------|--------|--------|------------|------| | [Role KPI 1] | | | | | | [Role KPI 2] | | | | | | On-time delivery rate | | | | | | Revisions per task | | | | | KPI summary: Met / Near / Not met — with the specific numbers. ## II. Strengths this period | Strength | Specific example (date + situation) | Impact on the team or KPI | |----------|-------------------------------------|---------------------------| ## III. Areas to improve | Item | Observed behavior | Actual impact | Specific expectation | |------|-------------------|---------------|----------------------| ## IV. Development plan 30-day goals: - [ ] [Specific, measurable skill or behavior] 90-day goals: - [ ] [...] Support from the leader: [training, materials, check-in frequency] ## V. Feedback from the team member - What do they think they contributed this period? - What support do they need from the leader? - What is blocking them that the leader does not know about? - Which direction do they want to grow? ## VI. Overall rating | Criterion | Score 1-5 | Comment | |-----------|-----------|---------| | Output quality | | | | On-time delivery | | | | Initiative | | | | Communication and collaboration | | | | Learning and improvement speed | | | | **Overall** | | | Scale: 1 = needs significant improvement · 3 = meets expectations · 5 = exceeds expectations. ## VII. Next steps - [ ] Send the document to the reviewee within 24 hours of the 1:1 - [ ] Set a 30-day follow-up reminder - [ ] Any compensation change: route through HR before it is stated in writing
| Duration | Content | |----------|---------| | 5 min | Ask for their self-assessment first | | 20 min | Walk the KPIs and data together | | 20 min | Strengths and areas to improve, listen to their response | | 10 min | Development plan — discuss until they agree, do not impose | | 5 min | Their feedback for the leader |
The leader prepares the document before the session and does not write it live during the meeting. For distributed teams, hold the 1:1 on video rather than in writing — written-only performance feedback reads harsher than intended and removes the chance to respond.
64-team-brief-global: source of on-time rate and revision counts.62-marketing-review-global and 47-design-review-global: approval outcomes serve as quantitative evidence.63-campaign-retrospective-global: the retro covers the system, this review covers the individual — never merge the two documents.07-marketing-report-global: results for the channels or campaigns they own.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 21,782 | 19,680 | -10% | 1 | 1 | 0% | 3,191 | 4,845 | +52% | 0 | 0 | — |
case-02 | fail→pass | 28,459 | 15,913 | -44% | 1 | 1 | 0% | 4,315 | 4,657 | +8% | 0 | 0 | — |
case-03 | fail→pass | 26,891 | 19,780 | -26% | 1 | 1 | 0% | 3,778 | 4,610 | +22% | 0 | 0 | — |
case-04 | pass→pass | 20,269 | 16,526 | -18% | 1 | 1 | 0% | 2,610 | 4,343 | +66% | 0 | 0 | — |
case-05 | pass→fail | 27,573 | 10,371 | -62% | 1 | 1 | 0% | 3,098 | 3,232 | +4% | 0 | 0 | — |
case-06 | pass→pass | 16,785 | 12,275 | -27% | 1 | 1 | 0% | 2,715 | 3,491 | +29% | 0 | 0 | — |
case-07 | fail→pass | 11,606 | 18,545 | +60% | 1 | 1 | 0% | 1,942 | 4,982 | +157% | 0 | 0 | — |
case-08 | fail→pass | 13,835 | 4,516 | -67% | 1 | 1 | 0% | 2,197 | 2,398 | +9% | 0 | 0 | — |
case-09 | fail→pass | 11,439 | 17,135 | +50% | 1 | 1 | 0% | 2,062 | 4,733 | +130% | 0 | 0 | — |
case-10 | fail→pass | 12,502 | 14,968 | +20% | 1 | 1 | 0% | 2,185 | 4,406 | +102% | 0 | 0 | — |
case-11 | pass→pass | 14,026 | 13,348 | -5% | 1 | 1 | 0% | 2,238 | 4,183 | +87% | 0 | 0 | — |
case-12 | fail→pass | 11,135 | 14,744 | +32% | 1 | 1 | 0% | 1,956 | 4,491 | +130% | 0 | 0 | — |
case-13 | pass→pass | 11,922 | 17,879 | +50% | 1 | 1 | 0% | 2,004 | 4,979 | +148% | 0 | 0 | — |
case-14 | fail→pass | 14,560 | 12,470 | -14% | 1 | 1 | 0% | 2,358 | 3,815 | +62% | 0 | 0 | — |
case-15 | fail→pass | 11,583 | 15,051 | +30% | 1 | 1 | 0% | 2,284 | 4,565 | +100% | 0 | 0 | — |
case-16 | fail→pass | 10,701 | 16,495 | +54% | 1 | 1 | 0% | 1,874 | 4,813 | +157% | 0 | 0 | — |
case-17 | fail→pass | 15,510 | 20,474 | +32% | 1 | 1 | 0% | 2,736 | 5,183 | +89% | 0 | 0 | — |
case-18 | fail→pass | 10,213 | 12,667 | +24% | 1 | 1 | 0% | 1,749 | 4,271 | +144% | 0 | 0 | — |
case-19 | fail→pass | 11,589 | 16,248 | +40% | 1 | 1 | 0% | 2,303 | 4,765 | +107% | 0 | 0 | — |
case-20 | fail→pass | 7,420 | 5,044 | -32% | 1 | 1 | 0% | 1,305 | 2,535 | +94% | 0 | 0 | — |
case-21 | fail→pass | 10,182 | 12,956 | +27% | 1 | 1 | 0% | 1,952 | 4,375 | +124% | 0 | 0 | — |
case-22 | pass→pass | 13,489 | 15,603 | +16% | 1 | 1 | 0% | 2,297 | 4,120 | +79% | 0 | 0 | — |
case-23 | pass→pass | 10,028 | 7,872 | -21% | 1 | 1 | 0% | 1,560 | 3,124 | +100% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted. The headline lift of +65 percentage points is the difference between those two pass rates over the 23 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.