Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Summarize a meeting transcript into structured notes with date, participants, topic, key decisions, summary points, and action items. Use when processing meeting recordings, creating meeting notes, writing meeting minutes, or recapping discussions.
.claude/skills/phuryn-summarize-meeting/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 44% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -2% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 57% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 118% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 198% | 0% |
You are an experienced product manager responsible for creating clear, actionable meeting summaries from $ARGUMENTS. This skill transforms raw meeting transcripts into structured, accessible summaries that keep teams aligned and accountable.
Meeting summaries are how knowledge spreads and accountability stays clear in product teams. A well-structured summary captures decisions, key points, and action items in language everyone can understand, regardless of who attended.
## Meeting Summary
Date & Time: Date and start/end time]
Participants: Full names and roles, if available]
Topic: Short title—what was the meeting about?]
Summary
Action Items
| Due Date | Owner | Action | |----------|-------|--------| | Date] | Name] | What needs to happen] | | Date] | Name] | What needs to happen] |
Decisions Made
Open Questions
Meeting-Summary-[date]-[topic].md| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 9,038 | 9,480 | +5% | 1 | 1 | 0% | 1,604 | 2,305 | +44% | 0 | 0 | — |
case-02 | fail→pass | 11,680 | 8,117 | -31% | 1 | 1 | 0% | 2,213 | 2,169 | -2% | 0 | 0 | — |
case-03 | pass→fail | 11,090 | 5,072 | -54% | 1 | 1 | 0% | 2,128 | 1,568 | -26% | 0 | 0 | — |
case-04 | fail→pass | 7,474 | 12,117 | +62% | 1 | 1 | 0% | 1,372 | 2,152 | +57% | 0 | 0 | — |
case-05 | fail→pass | 5,837 | 8,266 | +42% | 1 | 1 | 0% | 1,011 | 2,199 | +118% | 0 | 0 | — |
case-06 | pass→pass | 12,083 | 10,876 | -10% | 1 | 1 | 0% | 1,920 | 2,499 | +30% | 0 | 0 | — |
case-07 | pass→pass | 12,190 | 7,831 | -36% | 1 | 1 | 0% | 1,524 | 2,021 | +33% | 0 | 0 | — |
case-08 | fail→fail | 2,513 | 7,400 | +194% | 1 | 1 | 0% | 412 | 1,933 | +369% | 0 | 0 | — |
case-09 | fail→pass | 3,844 | 8,421 | +119% | 1 | 1 | 0% | 635 | 1,892 | +198% | 0 | 0 | — |
case-10 | fail→pass | 9,795 | 8,974 | -8% | 1 | 1 | 0% | 1,405 | 2,177 | +55% | 0 | 0 | — |
case-11 | fail→fail | 6,276 | 9,246 | +47% | 1 | 1 | 0% | 1,055 | 1,456 | +38% | 0 | 0 | — |
case-12 | fail→pass | 6,290 | 9,854 | +57% | 1 | 1 | 0% | 1,136 | 2,109 | +86% | 0 | 0 | — |
case-13 | pass→pass | 7,097 | 9,034 | +27% | 1 | 1 | 0% | 1,301 | 1,847 | +42% | 0 | 0 | — |
case-14 | fail→pass | 5,264 | 12,766 | +143% | 1 | 1 | 0% | 888 | 2,382 | +168% | 0 | 0 | — |
case-15 | pass→pass | 12,305 | 10,114 | -18% | 1 | 1 | 0% | 1,713 | 2,127 | +24% | 0 | 0 | — |
case-16 | fail→pass | 3,147 | 5,624 | +79% | 1 | 1 | 0% | 479 | 1,634 | +241% | 0 | 0 | — |
case-17 | fail→pass | 2,435 | 7,248 | +198% | 1 | 1 | 0% | 370 | 1,669 | +351% | 0 | 0 | — |
case-18 | pass→pass | 11,155 | 11,492 | +3% | 1 | 1 | 0% | 1,795 | 2,195 | +22% | 0 | 0 | — |
case-19 | fail→pass | 3,449 | 6,636 | +92% | 1 | 1 | 0% | 535 | 1,732 | +224% | 0 | 0 | — |
case-20 | pass→pass | 28,221 | 30,542 | +8% | 1 | 1 | 0% | 5,481 | 5,138 | -6% | 0 | 0 | — |
case-21 | pass→pass | 19,708 | 13,319 | -32% | 1 | 1 | 0% | 3,349 | 2,936 | -12% | 0 | 0 | — |
case-22 | pass→fail | 16,761 | 12,287 | -27% | 1 | 1 | 0% | 2,167 | 2,302 | +6% | 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. 22 cases were attempted. The headline lift of +41 percentage points is the difference between those two pass rates over the 22 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.