# Generate Ppt

> Produce brand-compliant PowerPoint deck via pptxgenjs, 4-gate quality pipeline.

- Skill: `jokerman89/generate-ppt` (Agent Skill)
- Install (CLI): `npx skillmds@latest add jokerman89/generate-ppt`
- Raw SKILL.md: https://api.skillmd.com/api/skills/jokerman89/generate-ppt/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: DevOps & Infra
- Author: jokerman89 (https://skillmd.com/u/jokerman89)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/jokerman89/generate-ppt

---


# /generate-ppt

Produces a brand-compliant PowerPoint deck (.pptx) for customer engagements. Uses pptxgenjs under the hood. Pulls templates from `~/.lintel/brand/ppt-templates/` (or falls back to in-repo defaults if brand not pulled). Compliance-gated before distribution.

## Prerequisites

This skill renders `.pptx` via the Node library [`pptxgenjs`](https://www.npmjs.com/package/pptxgenjs).
If it is not installed, set it up first: `npm i pptxgenjs` (project-local) or `npm i -g pptxgenjs`.
**Graceful degradation if it cannot be installed** (no Node toolchain / offline): fall back to
`/li:generate-web` → `/li:make-pdf` for a rendered deliverable, or emit the slide content as
markdown. Do not silently produce nothing — state which path you took.

## When to use

- Customer-engagement deliverable (pitch deck, workshop deck, summary deck)
- Internal presentation that needs the active pack's brand identity
- Reusing engagement-specific content across multiple PPT outputs
- Replacing manual PPT authoring with a workflow that has voice + brand + provenance gates

## When NOT to use

- One-slide quick mockup — `/design-html` is faster
- Non-PPT artifact — use `/generate-word` or `/generate-web`
- Pack compliance gate not satisfied AND output is customer-bound — resolve the gate first

## Inputs

- Required `--brief <path|inline>` — content brief describing the deck purpose **OR** `--from-pipeline <dir>` (shared pipeline mode)
- Required `--template <name>` — PPT template name from `~/.lintel/brand/ppt-templates/` (e.g. `pitch-deck`, `workshop`)
- Optional `--audience <text>` — primary audience (affects voice tier output)
- Optional `--slide-count <N>` — target slide count (default: 20-30 based on duration)
- Optional `--duration <minutes>` — presentation duration (informs slide pacing)
- Optional `--voice <tier>` — voice tier for slide content (default: the active pack's voice tier, `internal` by default)
- Optional `--use-defaults` — force use of in-repo default templates instead of brand pull
- Optional `--ignore-stale-brand <reason>` — bypass brand-staleness-warn

## From-pipeline mode (v3.5 Phase 2 — generate-pipeline integration)

If invoked with `--from-pipeline <run-dir>` instead of `--brief`:

1. **Read shared pipeline-output:**
   - `<run-dir>/content.md` — written content (with HTML-comment annotations for voice/type/key_message per section)
   - `<run-dir>/speaker-notes.md` — speaker notes per slide
   - `<run-dir>/design-spec.json` — per-format layout-mappings (read `per_format.ppt.layouts`)

2. **Replace brief-parsing logic** with direct-read of content.md sections + design-spec layouts. The PPTNarrativeArchitect agent is NOT invoked in this mode (narrative-design already done by shared generate-design).

3. **Apply format-specific design-pass via design_pass_hook:**
   - Reads `per_format.ppt.layouts[N].design_pass_hook` (canonical: PPTNarrativeArchitect)
   - Invokes agent for a PPT-specific fidelity-pass (slide-narrative-arc, FastPath-recommendations, animation-cues) on top of the shared baseline
   - Acknowledges plan-eng-review Reviewer Concern #7 — per-format design-agents stay per-format

4. **CLI stays backward-compat:** existing `--brief`-flag invocations work unchanged. `--from-pipeline` is additive.

5. **4-gate pipeline runs as usual** after generation (brand + voice + honest-limitations + provenance).

Operator-CLI stays non-breaking. Existing workflow-scripts are unaffected.

## Workflow

1. **Preflight gates:**
   - `~/.lintel/brand/ppt-templates/<template>.pptx` exists OR `--use-defaults` flag present
   - brand-staleness-warn check (90-day rule) — surface warning if stale
   - The active pack's compliance gates if a customer-facing voice tier is set (`resolve_pack_field compliance.hooks`; none by default — gate failure → output marked unverified)

2. **Read brief + extract structure:**
   - Goal / key message (one sentence)
   - Audience profile
   - 3-5 substantive sections
   - Opening hook + closing CTA

2b. **Design DNA slide pass (ADR-0017 — retrieval before slide design).** Query the slide
   decision engine so the arc is grounded in the corpus, not invented:
   ```bash
   dna="${LINTEL_SKILLS_DIR:-skills}/design-dna"
   python3 "$dna/scripts/search.py" "<deck goal / pitch type>" --slide strategy -n 1   # narrative arc + sparkline-beats
   # then per slide, by the slide's emotion (trust|urgency|confidence…) and goal (hook|proof|cta…):
   python3 "$dna/scripts/search.py" "<emotion>" --slide color-logic -n 1   # background/text/accent + full-bleed
   python3 "$dna/scripts/search.py" "<goal>"    --slide layout-logic -n 1  # layout pattern + break-pattern
   python3 "$dna/scripts/search.py" "<slide-type>" --slide copy -n 1       # headline formula
   ```
   Feed the strategy's `sparkline_beats` + `emotion_arc` to PPTNarrativeArchitect as the arc spine.
   python3 absent → Read `design-dna/data/slides/*.csv` directly (controlled vocabulary in design-dna SKILL.md).

3. **Invoke `PPTNarrativeArchitect` agent** to design slide arc:
   - Slide-by-slide content goals — seeded by the slide-strategy `sparkline_beats`
   - Mode tags per slide (Reveal / Inspire / Provoke / Neutral) — aligned to the slide's emotion
   - Layout suggestions per slide — from `--slide layout-logic` (pattern + break-pattern at 1/3, 2/3)
   - Asset suggestions (from the active pack's asset library, if one is configured)

4. **Generate slides via pptxgenjs:**
   - Apply template
   - Add slides per architect's arc
   - Populate text with brief-derived content
   - Insert assets from the active pack's asset library where matched
   - Mode-tag in slide notes for voice-check downstream

4b. **Critique slide narration via `SlideNarrationCritic` agent** (if the deck carries speaker notes / a talk track):

```bash
narration_brief=$(mktemp)
cat > "$narration_brief" <<EOF
task: Critique the generated deck's slide narration / speaker notes
context_pointers:
  - <draft deck path>.pptx (speaker notes per slide)
constraints:
  - voice-tier consistency per active pack
  - pacing (words-per-minute vs each slide's duration)
  - audience alignment + recovery-line presence on risky slides
acceptance:
  - per-slide narration verdict + specific rewrite recommendations
EOF

/li:brief-forge subagent_spawn generate-ppt SlideNarrationCritic brief "$narration_brief"
```

5. **4-gate quality pipeline** (output stays in `~/.lintel/draft/` until ALL 4 PASS):

   **Gate 1 — Voice (the active pack's compliance gates):**
   - Apply the active pack's voice rubric to slide text (`resolve_pack_field compliance.hooks`; none by default)
   - Require any pack-defined threshold met, 0 P1 violations
   - PASS → continue; FAIL → surface + regenerate flagged slides (up to 2 retries)

   **Gate 2 — Brand-conformance:**
   - Template matches latest brand version (or default-fallback marker present)
   - Assets are from the active pack's asset library (or operator-confirmed)
   - No unauthorized branding (third-party logos)
   - Color/typography matches template tokens

   **Gate 3 — Honest-limitations:**
   - Applies to transparency-note slides ONLY
   - Limitations section ≥ capabilities − 2 (per the active pack's disclosure rule, if defined)
   - Skipped for pitch/workshop decks (no transparency-note section)

   **Gate 4 — Provenance:**
   - Generate a provenance record with source chain + voice score + brand version
   - Record landed in `.claude/runtime/audit/`

6. **On all 4 PASS:** move from `~/.lintel/draft/` → operator-specified `--out` path (or `<brief-stem>.pptx` in cwd).

7. **On any gate FAIL:** keep in draft, surface specific failures, allow operator iteration.

## Report format

```
Generate PPT: customer-A-arc-pitch

Template: pitch-deck.pptx (~/.lintel/brand/ppt-templates/, brand version 2026-Q2)
Brand staleness: ok (32 days)
Voice tier: internal (pack-resolved)
Audience: mid-market public sector IT leadership

## Slide architecture (PPTNarrativeArchitect)
  1. Opening: Reveal/Curtain — "What does Friday 2 AM actually cost?"
  2. Stakes: Provoke/Unflinching — current cost in incidents + hours
  3. Setup: Reveal/Dream — picture the unified-Arc world
  4-8. Substantive: feature surface mapped to operational pain
  9. Close: Inspire/Marvel — what becomes possible

## Generation (pptxgenjs)
  9 slides generated. 6 service icons resolved from the active pack's asset library.
  142 KB output → ~/.lintel/draft/customer-A-arc-pitch.pptx

## 4-Gate pipeline
  Gate 1 (voice):   ✓ PASS — score 87/100, 0 P1 violations
  Gate 2 (brand):   ✓ PASS — template + 6/6 assets from brand-version 2026-Q2
  Gate 3 (honest):  N/A — no transparency-note slides
  Gate 4 (proven):  ✓ PASS — PROV-7f8a2 recorded

## Status
ALL GATES PASS. Moving from draft → ./customer-A-arc-pitch.pptx.

Distribution: operator-driven. Verify the provenance chain (PROV-7f8a2) before sending.
```

## Compliance integration

- 4-gate pipeline is the customer-bound enforcement path; the specific gates are pack-configurable (`resolve_pack_field compliance.hooks`; none by default)
- The active pack's deploy/release gate (if any) reads provenance + voice-check status before customer-bearing distribution
- Customer-data patterns in brief → BLOCK (always-on)
- Stale brand (>90d) surfaces warn but doesn't block (operator decides)

## Voice tier note

`voice: mixed`. Skill itself is engineering-internal; the SLIDE CONTENT may use a customer-facing voice tier (gated by Gate 1).

## Failure modes

- **pptxgenjs runtime error** (lib bug, malformed template) — surface error, keep work-in-progress in `~/.lintel/draft/.work/`, allow operator manual debug
- **Brand template missing AND --use-defaults not set** — surface options: pull brand, use defaults, abort
- **Voice gate fails after 2 regen attempts** — keep draft, surface specific slide failures with fix recommendations
- **Asset library returns 0 results for slide concept** — surface to operator, allow them to provide path manually OR skip the asset for that slide
- **Customer-data in brief** — BLOCK before generation. Sanitize first.

## Examples

**Standard pitch deck:**
```
> /generate-ppt --brief docs/engagement/customer-A-pitch-brief.md --template pitch-deck --audience "public-sector CIO"
[Architect designs, pptxgenjs generates, 4 gates pass]
✓ Deck at ./customer-A-arc-pitch.pptx. Provenance PROV-7f8a2.
```

**Workshop deck with defaults:**
```
> /generate-ppt --brief workshop-brief.md --template workshop --use-defaults --voice internal
[Uses in-repo default-ppt-template; voice tier internal so Gate 1 is no-op]
✓ Deck at ./workshop.pptx (default-fallback marker present).
```

**Stale brand override:**
```
> /generate-ppt --brief archive-pitch.md --template pitch-deck --ignore-stale-brand "archival deliverable, brand version pinned"
[Bypasses brand-staleness-warn, logs reason]
```

## See also

- `BRAND-INTEGRATION.md` — brand architecture + cache
- `PPTNarrativeArchitect` agent — slide arc design
- `SlideNarrationCritic` agent — slide narration / talk-track critique (post-generation)
- The active pack's compliance gates — Gate 1
- `/generate-word`, `/generate-web` — sibling doc-gen skills
- `brand-staleness-warn` hook — Phase E