Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Turn a vague concern into specific, kind, actionable feedback. Use when asked to give feedback, write a feedback note, prepare to tell someone something hard about their work, or coach a report/peer. Produces ready-to-deliver feedback structured on situation–behaviour–impact, separating observation from judgement, with the change requested and an opening line — calibrated to praise or constructive.
.claude/skills/mohitagw15856-giving-feedback/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-13 | ✗→✓ | ▲ Improved | 120% | 0% |
| case-04 | ✓→✗ | ▼ Worse | 49% | 0% |
| case-07 | ✓→✗ | ▼ Worse | 25% | 0% |
| case-21 | ✓→✓ | ＝ Same ✓ | 78% | 0% |
| case-11 | ✓→✓ | ＝ Same ✓ | 40% | 0% |
Most feedback is useless because it's vague ("be more proactive"), judgemental ("you're careless"), or sandwiched into mush. Good feedback is specific, describes behaviour not character, names the impact, and makes the ask clear. This skill turns a fuzzy concern into feedback the person can actually act on — delivered with enough care that they hear it.
Ask for these only if they aren't already provided:
1. The core (SBI) — the spine of good feedback:
2. The ask — for constructive: the specific change ("next time, walk the pricing slide before Q&A"). For praise: name what to keep doing and why it mattered (praise that's specific gets repeated).
3. Opening line — how to start so they're ready to hear it (ask permission / state intent: "Can I share something from the review? I want the next one to land even better.").
4. Make it a dialogue — 1–2 questions to invite their view ("How did it feel from your side?"), because feedback is a conversation, not a verdict.
Calibration note — keep it timely (soon, not saved for the review), private if constructive, and about the behaviour, never the person.
SBI feedback model (Center for Creative Leadership) and Radical Candor (Kim Scott) — care personally, challenge directly.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-21 | pass→pass | 6,780 | 9,069 | +34% | 1 | 1 | 0% | 1,273 | 2,267 | +78% | 0 | 0 | — |
case-11 | pass→pass | 10,579 | 10,058 | -5% | 1 | 1 | 0% | 1,628 | 2,285 | +40% | 0 | 0 | — |
case-01 | pass→pass | 8,705 | 6,730 | -23% | 1 | 1 | 0% | 1,558 | 1,994 | +28% | 0 | 0 | — |
case-02 | pass→pass | 8,714 | 6,172 | -29% | 1 | 1 | 0% | 1,507 | 1,764 | +17% | 0 | 0 | — |
case-03 | pass→pass | 8,755 | 5,733 | -35% | 1 | 1 | 0% | 1,593 | 1,879 | +18% | 0 | 0 | — |
case-04 | pass→fail | 11,256 | 14,287 | +27% | 1 | 1 | 0% | 2,085 | 3,116 | +49% | 0 | 0 | — |
case-05 | pass→pass | 14,734 | 15,956 | +8% | 1 | 1 | 0% | 2,389 | 3,321 | +39% | 0 | 0 | — |
case-06 | pass→pass | 12,828 | 14,182 | +11% | 1 | 1 | 0% | 2,372 | 3,085 | +30% | 0 | 0 | — |
case-07 | pass→fail | 9,662 | 6,657 | -31% | 1 | 1 | 0% | 1,572 | 1,958 | +25% | 0 | 0 | — |
case-08 | pass→pass | 12,014 | 8,767 | -27% | 1 | 1 | 0% | 1,876 | 2,313 | +23% | 0 | 0 | — |
case-09 | pass→pass | 11,424 | 9,925 | -13% | 1 | 1 | 0% | 1,856 | 2,346 | +26% | 0 | 0 | — |
case-10 | pass→pass | 5,944 | 6,329 | +6% | 1 | 1 | 0% | 1,099 | 1,870 | +70% | 0 | 0 | — |
case-12 | pass→pass | 9,066 | 8,432 | -7% | 1 | 1 | 0% | 1,501 | 2,068 | +38% | 0 | 0 | — |
case-13 | fail→pass | 6,157 | 8,693 | +41% | 1 | 1 | 0% | 1,004 | 2,208 | +120% | 0 | 0 | — |
case-14 | pass→pass | 10,339 | 7,551 | -27% | 1 | 1 | 0% | 1,688 | 1,874 | +11% | 0 | 0 | — |
case-15 | pass→pass | 8,034 | 5,081 | -37% | 1 | 1 | 0% | 1,399 | 1,645 | +18% | 0 | 0 | — |
case-16 | pass→pass | 11,899 | 7,883 | -34% | 1 | 1 | 0% | 1,894 | 1,935 | +2% | 0 | 0 | — |
case-17 | pass→pass | 7,598 | 8,441 | +11% | 1 | 1 | 0% | 1,196 | 2,059 | +72% | 0 | 0 | — |
case-18 | pass→pass | 10,347 | 9,911 | -4% | 1 | 1 | 0% | 1,826 | 2,379 | +30% | 0 | 0 | — |
case-19 | pass→pass | 6,143 | 6,001 | -2% | 1 | 1 | 0% | 993 | 1,706 | +72% | 0 | 0 | — |
case-20 | pass→pass | 7,759 | 7,226 | -7% | 1 | 1 | 0% | 1,282 | 1,928 | +50% | 0 | 0 | — |
case-22 | pass→pass | 10,835 | 9,965 | -8% | 1 | 1 | 0% | 1,799 | 2,254 | +25% | 0 | 0 | — |
case-23 | pass→pass | 8,258 | 6,287 | -24% | 1 | 1 | 0% | 1,499 | 1,793 | +20% | 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 -33 percentage points is the difference between those two pass rates over the 23 comparable cases. 2 cases got worse with the skill loaded, and they are 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.