presentation-creator

Bold, minimal slide decks with a story underneath: spine to final QA.

• IS: slide decks end to end: story spine, narrative structure, slide copy, visual design, speaker notes, investor pitch decks, and decks built as a web app.
• IS NOT: long-form prose (use the external ghostwriter skill with platform blog), marketing copy outside slides (use copywriting), or general visual systems for product UI (use ui-design).

Workflow

Track this checklist:

Presentation progress:
- [ ] Step 1: Gather context (audience, setting, format, output)
- [ ] Step 2: Write the story spine and the ending (references/story-structure.md)
- [ ] Step 3: Outline the slide sequence (references/outline-structure.md)
- [ ] Step 4: Write slide copy (references/writing-slides.md; pitch decks: references/pitch-decks.md instead)
- [ ] Step 5: Design visual layout and composition (references/visual-design.md)
- [ ] Step 6: Write speaker notes (references/speaker-notes.md); skip for pitch decks
- [ ] Step 7: QA pass, output the slide-by-slide review table

Step 1: Gather context

Establish four things (ask if not provided):

• Audience: internal (shared context, be direct) vs. external (build credibility, define terms)
• Setting: live talk, recorded/async, or standalone investor pitch
• The three messages: what the audience must remember after the deck
• The output: a markdown outline to hand to a deck tool, a deck built as a web app, or slides in an existing template. Ask before Step 5, because the visual step is where it starts to matter

Route by format:

• Live talk or internal/recorded deck → Steps 2-7 in order.
• Investor pitch deck (read without a presenter) → read references/pitch-decks.md first. Its 10-slide framework replaces Step 3’s outline, and its async copy rules (denser, standalone headlines, 2-3 bullets per section) replace writing-slides.md for Step 4: do not load writing-slides.md on this path, its presented-deck copy limits contradict it. Step 2 still applies, in a compressed form: an investor is an audience too, and the spine is what stops a pitch reading as a feature list. Skip Step 6 (no presenter); Steps 5 and 7 still apply.

Steps 2-6: Build the deck

Read each step’s reference when you reach it:

Building the deck as a web app rather than exporting from a deck tool: references/web-deck.md covers the route-per-slide structure, navigation, layout primitives, motion, and live demos. Read it at Step 5, after the copy exists, never before Step 3.

Step 7: QA pass (produces evidence)

Review every slide and output a table. This is the deliverable that proves the deck is done, not a “looks good” sign-off:

| # | Slide | 3-sec test | One message | Spine beat | Layout | Color |
|---|-------|-----------|-------------|------------|--------|-------|
| 1 | Title | pass | pass | once upon a time | full statement | teal |

• 3-sec test: parseable in 3 seconds at arm’s length? Cut copy on any failure until it passes.
• One message: exactly one idea per slide; split slides carrying two.
• Spine beat: which beat of the Step 2 spine this slide serves. A slide serving none is a fact you found interesting; cut it.
• Layout: no more than 2 consecutive slides with the same layout.
• Color: the section accent, or the full-bleed palette, matching what Step 5 assigned.

Deck-level checks below the table:

• Every spine beat has at least one slide, and the ending matches the one written first
• The deck states a position a reasonable person could disagree with
• One colour system throughout: accents per section, or full-bleed palettes, never both
• Recap slide has exactly one line per core section
• Pitch decks only: ≤15 slides, explicit ask slide (amount + use of funds), headlines pass the forwardable test (make sense with zero context)

Fix every flagged row and re-output the table before handing over.

Core principles

• Story before slides: the spine decides which slides exist. Write the ending first
• Take a position: a deck nobody could disagree with has not said anything
• Headlines do the work: bold statements, not topic labels:

  Before: "An Overview of Our Q3 Performance Metrics and Results"
  After:  "Q3: Revenue Up 40%. Here's How."

• Impact through scale, not weight: large light type beats small bold type
• One colour system, held for the whole deck: full-bleed palettes where a palette owns the entire slide, or dark with one accent per section. Either creates the rhythm the audience tracks position by
• Demo it live where you can: a working demo, not a screenshot of one

Gotchas

• A deck with no opinion: surveying the landscape feels safe to write and gives the room nothing to hold. If nobody could disagree, there is no talk yet.
• Paragraphs on slides: the audience reads instead of listening and the speaker becomes redundant. Fail the 3-second arm’s-length test, cut copy until it passes.
• Accents outside the section system: section colors are wayfinding; a random mid-section accent reads as a topic change that never happened. One color per major section, teal reused for opening and closing.
• Accent thinking on a full-bleed deck: adding a highlight colour to a slide whose palette already owns the screen. On that system the slide is the accent; the colour changes at the slide boundary, not inside it.
• Fixed pixel type: a deck sized for the presenter’s laptop is a different deck on the projector and unreadable on the phone it gets forwarded to. Size in clamp().
• A screenshot where a live demo would land: if the deck is a web app and the thing runs, run it. A screenshot of a working demo is the weaker version of both.
• Speaker notes as a script: a verbatim script gets read aloud and sounds flat. Notes are scannable prompts: key point, talk-track bullets, transition line.
• Same layout on every slide: uniform layouts flatten rhythm; the audience stops registering new slides. Alternate full-statement, split, and data layouts per visual-design.md.
• Skipping the spine: jumping straight to slides produces a list of facts with no arc, then a rewrite once the missing narrative shows. Spine and ending first, then the outline.
• Building the deck app before the story exists: the primitives get designed around slide 3 and fight every slide after it.
• Sparse headlines on pitch decks: “Traction” tells a skimming investor nothing. Write the complete claim: “1,000+ Customers, $10M ARR”. Pitch decks are read, not presented.
• Presented-deck density on a pitch deck: a 3-words-per-slide deck forwarded with no presenter is unreadable. Route to pitch-decks.md in Step 1, not after the deck is built.

Related skills

• Optional external ghostwriter where installed: long-form articles and tutorials from the blog platform profile; use when the output is prose, not slides
• copywriting: landing pages, CTAs, marketing copy outside a deck
• ui-design: visual systems for product UI and landing pages; presentation visual rules live in references/visual-design.md instead
• Taste Training (blode.co/taste-training): trains the eye these rules encode, across type, copy, craft, interaction, and motion