/create-pitch-deck

name: create-pitch-deck
description: Draft a 10-slide pitch deck for a Sui product (investors, judges, grant). Use when the user wants a pitch deck or to draft slides for a Sui project.

Preamble (run first)

# Suiperpower telemetry. Routes through the CLI so projects.json + Convex stay in sync.
# Silent on failure so the skill never blocks. Opt out: SUIPERPOWER_TELEMETRY=off.
#
# AGENT NOTE: when this skill finishes, run the matching completion command:
#   suiperpower track create-pitch-deck ship completed
# Or use "failed" / "aborted" if it ended that way.
command -v suiperpower >/dev/null 2>&1 && suiperpower track create-pitch-deck ship started >/dev/null 2>&1 &
true

If `TEL_PROMPTED` is `no`, before doing real work, ask the user:

> Help suiperpower get better. We track which skills get used and how long they take. No code, no file paths, no PII. Change anytime in `~/.suiperpower/config.json`. > > A) Sure, anonymous > B) No thanks

Write the answer to `~/.suiperpower/config.json` `telemetryTier` field and create `~/.suiperpower/.telemetry-prompted`. Then continue.

What this skill does

Drafts a 10-slide pitch deck outline plus per-slide copy, grounded strictly in the project's context files. Picks a structure depending on the audience (hackathon judges, investors, or grant program reviewers). Refuses to claim any metric that the source files do not support.

The output is a written deck plan in `.suiperpower/pitch-deck.md` that the user can render in any deck tool (Keynote, Pitch, Figma, Tome). The skill does not produce slide images.

When to use it

• Before submitting to Sui Overflow alongside `submit-to-sui-overflow`.
• For a Sui Foundation grant application alongside `apply-grant`.
• For an investor meeting where the user has 10 slides to make a case.

When NOT to use it

• When the project has no validated context yet. Run `find-next-sui-idea` and `validate-business-model` first.
• For a one-page summary or a leave-behind document. Different format.
• For marketing copy aimed at end users. Route to a marketing skill.

If you activated this and the user actually wants something else, consult `skills/SKILL_ROUTER.md` and hand off.

Inputs

• The audience: `judges`, `investors`, or `grant`. The structure changes per audience.
• `.suiperpower/idea-context.md`, `.suiperpower/business-model.md`, `.suiperpower/retention-loop.md`, `.suiperpower/deploy-context.md`, `.suiperpower/track-pick.md` if present.
• Optional: a recent `roast.md` or `product-review.md` so the deck addresses known weaknesses.

Outputs

A `.suiperpower/pitch-deck.md` with a 10-slide outline. Each slide has a title, a one-line claim, and the copy or visual reference per the audience structure:

## Pitch deck, <timestamp>

### Audience
- type: <judges | investors | grant>
- length target: <minutes for live pitch, or read-time for leave-behind>

### Slides

#### Slide 1, <title>
- claim: <one sentence>
- copy: <on-slide text>
- visual: <what to show, optional>

#### Slide 2, <title>
- ...

(through slide 10)

### Backup slides (if relevant)
- <appendix slide 1>
- <appendix slide 2>

### Notes
- <facts cited from context files>
- <claims that needed downgrading because the source did not support them>

Workflow

1. **Pick the audience structure**

• Judges: hook, problem, solution, demo, why-Sui, traction (if any), team, business, ask, close.
• Investors: hook, market, problem, solution, traction, business model, why-now, team, ask, close.
• Grant: hook, problem, solution, why-this-team, deliverables, milestones, budget, public-good rationale, sustainability, ask.
• Reference `references/deck-structures.md` for the per-slide rules.

2. **Read context files**

• Pull problem statement, target user, solution from `idea-context.md`.
• Pull business model from `business-model.md`. If the file is missing or the verdict is `no`, do not claim a model on the deck; the slide acknowledges it as open.
• Pull retention loop from `retention-loop.md`. Same rule.
• Pull package id and load-bearing track from `deploy-context.md` and `track-pick.md`.

3. **Draft each slide**

• One claim per slide. The claim must be a single sentence that survives outside the deck.
• On-slide copy is short. The narrator carries the rest.
• Visuals are described in words ("screenshot of demo flow at the load-bearing step", "table of three milestones with dates"). The skill does not produce images.

4. **Apply the no-fake-metrics rule**

• Any number on the deck must be cited to a source: a measurement, a public benchmark, or a committed forecast labeled as a forecast.
• If a number cannot be cited, either remove it or rewrite the claim qualitatively.
• Track every downgrade in the `Notes` section so the user knows what was rejected.

5. **Stress-test against the audience**

• Judges deck: does the demo slide actually carry the load-bearing flow?
• Investors deck: does the why-now slide name a recent change in market or tech?
• Grant deck: does the public-good rationale name what the user gets that they could not get from a closed equivalent?

6. **Writeback**

• Append `.suiperpower/pitch-deck.md` with the timestamped block.

7. **Hand off**

• Recommend `roast-my-product` if the user has not yet stress-tested the deck.
• Recommend `marketing-video` once the deck is done; the deck and video should share the same one-liner.

Quality gate (anti-slop)

Before reporting done:

• Did the audience get picked explicitly, and the right structure applied?
• Was every metric cited or removed? (No "millions of users" without a source.)
• Did the claim sentence on every slide survive outside the deck?
• Did the demo slide name the load-bearing flow, not a feature list?
• Did the deck avoid banned words and em-dashes?
• Did the writeback happen?

If any answer is no, the skill keeps working.

References

On-demand references (load when relevant to t