Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Turn the user's REAL doc into a slide deck — open the source, structure the narrative, and build the actual .pptx — not slide-writing tips. Use when asked to make a deck from this doc, turn my brief into slides, build the presentation from my Drive doc, or deckify this in Cowork. Reads the document via the Google Drive/Docs connector, maps it to a one-idea-per-slide narrative, and produces a real presentation artifact (.pptx) with speaker notes plus a slide-by-slide outline.
.claude/skills/mohitagw15856-deck-from-doc/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | 82% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -66% | 0% |
| case-11 | ✓→✗ | ▼ Worse | -36% | 0% |
| case-17 | ✓→✗ | ▼ Worse | -38% | 0% |
| case-04 | ✓→✓ | = Same ✓ | 20% | 0% |
A good doc isn't a good deck — prose has to become one idea per slide with a spine an audience can follow. In Claude Cowork this skill reads the real document and builds an actual presentation file, so the user gets slides to refine, not advice on how to make them.
.pptx (or Slides), one idea per slide, with a clear narrative arc and speaker notesAsk for these if not provided:
pptx skill / presentation tooling in the sandbox to generate a real .pptx: title, one-idea slides with conclusion headlines, a data slide where the doc has numbers, a recommendation and an ask slide. Apply the brand template if provided.Guardrails: every slide claim traces to the doc — never invent data or a recommendation the doc doesn't support; keep numbers exact; if the connector/template is unavailable, build from what's given and note the fallback.
> one line
| # | Headline (takeaway) | The one point | |---|---|---|
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 30,034 | 27,934 | -7% | 1 | 1 | 0% | 6,217 | 7,055 | +13% | 0 | 0 | — |
case-02 | fail→fail | 31,068 | 33,978 | +9% | 1 | 1 | 0% | 6,211 | 7,050 | +14% | 0 | 0 | — |
case-03 | fail→fail | 5,883 | 4,596 | -22% | 1 | 1 | 0% | 956 | 1,523 | +59% | 0 | 0 | — |
case-04 | pass→pass | 14,680 | 12,530 | -15% | 1 | 1 | 0% | 2,210 | 2,653 | +20% | 0 | 0 | — |
case-05 | pass→pass | 24,464 | 22,965 | -6% | 1 | 1 | 0% | 5,458 | 5,804 | +6% | 0 | 0 | — |
case-06 | fail→pass | 3,813 | 4,651 | +22% | 1 | 1 | 0% | 892 | 1,620 | +82% | 0 | 0 | — |
case-07 | fail→fail | 9,202 | 3,730 | -59% | 1 | 1 | 0% | 1,743 | 1,506 | -14% | 0 | 0 | — |
case-08 | fail→fail | 16,968 | 25,406 | +50% | 1 | 1 | 0% | 2,797 | 6,416 | +129% | 0 | 0 | — |
case-09 | fail→pass | 27,054 | 7,525 | -72% | 1 | 1 | 0% | 6,187 | 2,129 | -66% | 0 | 0 | — |
case-10 | fail→fail | 8,025 | 4,607 | -43% | 1 | 1 | 0% | 1,384 | 1,686 | +22% | 0 | 0 | — |
case-11 | pass→fail | 12,646 | 4,131 | -67% | 1 | 1 | 0% | 2,450 | 1,579 | -36% | 0 | 0 | — |
case-12 | fail→fail | 2,000 | 8,205 | +310% | 1 | 1 | 0% | 359 | 2,189 | +510% | 0 | 0 | — |
case-13 | fail→fail | 26,824 | 34,160 | +27% | 1 | 1 | 0% | 4,924 | 7,014 | +42% | 0 | 0 | — |
case-14 | pass→pass | 8,757 | 4,864 | -44% | 1 | 1 | 0% | 1,533 | 1,765 | +15% | 0 | 0 | — |
case-15 | fail→fail | 10,341 | 20,268 | +96% | 1 | 1 | 0% | 2,010 | 4,437 | +121% | 0 | 0 | — |
case-16 | fail→fail | 14,387 | 27,590 | +92% | 1 | 1 | 0% | 2,740 | 7,016 | +156% | 0 | 0 | — |
case-17 | pass→fail | 15,165 | 4,821 | -68% | 1 | 1 | 0% | 2,759 | 1,712 | -38% | 0 | 0 | — |
case-18 | fail→fail | 13,178 | 3,903 | -70% | 1 | 1 | 0% | 2,547 | 1,537 | -40% | 0 | 0 | — |
case-19 | fail→fail | 3,599 | 22,478 | +525% | 1 | 1 | 0% | 658 | 5,337 | +711% | 0 | 0 | — |
case-20 | fail→fail | 2,211 | 3,677 | +66% | 1 | 1 | 0% | 372 | 1,426 | +283% | 0 | 0 | — |
case-21 | fail→fail | 9,065 | 5,356 | -41% | 1 | 1 | 0% | 1,593 | 1,677 | +5% | 0 | 0 | — |
case-22 | fail→fail | 10,587 | 5,155 | -51% | 1 | 1 | 0% | 1,708 | 1,807 | +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 0 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.