- Home
- Skills
- Image & Video
- slide deck designer
More screenshots
Works with the AI Agents you already use
slide deck designer
Your manager calls it "a database dump" and the decision you needed never happens.
$9.99
slide deck designer
Example session with this skill installed
I have a 12-slide Q4 review deck for our exec staff meeting Thursday and it's a disaster — my manager said it 'reads like a database dump'. Every slide is titled things like 'Overview' and 'Metrics', bullets everywhere, and the actual news (we cut onboarding churn 40% and need 2 QA hires approved) is buried on slide 9. 20 minutes slot, will also be forwarded to the board. Can you fix it?
- Read your context and instructions
- Compiled the slide deck designer
Verdict: Diagnosis confirmed — classic document-deck. The real story is strong; the deck hides it. Restructured to 8 slides, headlines rewritten as assertions, density cut 71%.
Headline walk-through (the new deck in 8 slides):
| # | New assertion headline | Was |
|---|---|---|
| 1 | "Q4: churn down 40%, one hire request stands between us and scale" | "Overview" |
| 2 | "Onboarding revamp cut churn 40% — from 12% to 7.2% in one quarter" | "Metrics" |
| 3 | "The win traces to three changes: checklist, concierge, rollback" | "Initiatives" |
| 4 | "Revenue held flat while churn fell — retention is compounding" | "Revenue" |
| 5 | "What is: QA is the bottleneck — 14-day median review cycle" | "QA Update" |
| 6 | "What could be: two QA hires cut the cycle to 5 days by March" | "Hiring" |
| 7 | "The ask: approve 2 QA headcount ($228K/yr) to protect the churn win" | "Next Steps" |
| 8 | Appendix: full metrics, methodology, cohort detail | — |
Read only the headlines in order — they already read as an executive summary. 12 slides → 8.
Density cut: slide 2 carried 118 words → now 24 words: one chart (final data point highlighted, rest grayed), annotation directly on the drop. Total deck: 1,140 → 330 words (-71%).
QA pass caught: silent Times font fallback (now explicit), slide-number blue links (styled), "compounding" claim flagged {{VERIFY}} instead of fabricated. Closing rewritten from "Next Steps" to the ask with decision deadline.
Delivered: draft HTML + storyboard + placeholder list + speaker notes on all 8 slides.
Connects securely to your tools. The creator never sees your data.
About this skill
The problem Ask an AI for "a deck about X" and you get the document-deck: it starts writing slide 1 with no story, labels every headline ("Overview", "Results", "Next Steps"), packs 120 words onto slides that compete with the speaker, sprays five chart styles and a rainbow palette, and ships in Times New Roman without noticing — because reveal.js scales overflowing slides silently and nobody re-checked at presentation size. The deck reads fine in code and dies in the room. Your manager calls it "a database dump" and the decision you needed never happens. What it does
- Gates the story before any slide exists — picks the narrative arc (Problem→Solution, SCQA opening, Rule of Threes) and outlines it before design starts
- Rewrites headlines as assertions — "Q3 Results" becomes "Q3 revenue up 23%, driven by mid-market"; headlines alone read as an executive summary
- Enforces word budgets per slide type — evidence slides ≤40 words, 120-word hard ceiling, density comes from structure not shrinking fonts
- Commits to a disciplined visual system — one canvas (dark or light, never blended), two primaries, one idea per slide, balanced rows, charts annotated directly on the data
- Runs the layout QA pass and ships with a delivery ledger — catches silent font fallback, gradient clipping, PDF background loss, placeholder list included, never fabricates statistics Why this beats prompting it yourself A generic prompt optimizes for "sounds done": it fills slides, it doesn't land stories. It will never tell you your headline labels instead of asserts, that slide 9 buries your only real news, that your font silently fell back to Times, or that your dark background vanishes in PDF export. Those failures are invisible in code review and fatal in the room — and they're exactly what the storyboard gate, word-budget audit, and QA checklist exist to catch. Use cases
- Pitch or sales deck where skimming investors read only the headlines
- Turning a dense report into a presentable narrative without bullet-point anesthesia
- Teaching/workshop deck with speaker notes that survive being forwarded
- Fixing an existing deck: storyboard diagnosis, headline rewrite, density cuts, QA pass Known limitations
- Produces HTML decks (reveal.js-compatible), not .pptx — request pptx only if explicitly needed and expect styling loss
- Does not generate photography or illustrated artwork; uses color, type, inline SVG, and your images
- Does not fabricate data — unsourced figures become visible placeholders for you to fill
- Not a BI tool — complex multi-series analytics should be exported as images by you What this skill refuses to do
- Write slides before the story outline exists
- Label headlines instead of asserting conclusions
- Ship a deck with fabricated statistics or invented customer quotes
- Blend dark and light canvas styles in one deck, or exceed the two-primary palette
- End a deck on "Thank you" when your goal was a decision
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Video Skills.
Outputs reveal.js HTML, not native .pptx. Cannot generate photographic assets or fabricate missing data points. High-complexity charts require manual image imports.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 5 days ago