# Deck Outline

> This skill should be used when the user asks to "outline a deck on X," "what's the storyline for this deck," "build a slide-by-slide outline," or needs a presentation structured before any actual slide design happens.

- Skill: `daddia/deck-outline` (Agent Skill)
- Install (CLI): `npx skillmds@latest add daddia/deck-outline`
- Raw SKILL.md: https://api.skillmd.com/api/skills/daddia/deck-outline/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Docs & Writing
- Author: daddia (https://skillmd.com/u/daddia)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/daddia/deck-outline

---

# Deck Outline

## When to use

For consultants building a storyline-first, action-titled deck outline before slide design. Markdown outline only — hand off to presentation tools after approval.

## What this skill does not do

- **Does not build slides** — outline only; pptx or design tools after approval.
- **Does not invent narrative** without governing thought — route to `narrative-builder` if needed.
- **Does not rewrite source analysis** — maps pyramid to slides.

## Preconditions

| Input | If missing |
|---|---|
| Practice profile (slide title style, exhibits) | Proceed with action titles default; flag practice-setup |
| Governing thought / narrative | Ask to run `narrative-builder` first |

## Provisional mode

Thin narrative: label structured first pass; flag slides that don't trace to key line.

## Trust spine

- **Confidence bands** (`narrative-synthesis`): High = every slide traces to key line; Medium = some exhibit TBD; Low = topic list without governing thought — flag.
- **Analytical Rigor:** Every slide supports a key line point; cut or flag non-load-bearing slides.
- **Escalation:** Slide without key-line parent → flag or cut; board-final → gate + reviewer note.

Full rules: `../../references/trust-conventions.md`.

## Workflow

1. Read practice profile for slide title style and exhibit conventions.
2. Use `narrative-builder` output if available; else ask to build narrative first.
3. Map pyramid: governing thought slide → key line slides → supporting evidence (one exhibit per slide).
4. Action titles unless profile says descriptive.
5. Note exhibit per slide without building it.
6. So-what test per `minto-pyramid.md` — cut non-load-bearing slides.

## Output format

```
GOVERNING THOUGHT: [headline]

Slide 1 — [Action title]
  Exhibit: [...]
Slide 2 — [Action title for key line point 1]
  Exhibit: [...]
...
```

Flag slides that don't trace to a key line point.

## Worked example

**Input:** Governing thought: "Mid-tier price increase destroyed volume — revert pricing." Three MECE key lines.

**Excerpt:** Slide 1 action title states governing thought; Slides 2–4 map one key line each with exhibit notes (cohort chart, elasticity model, competitor scan).

## Quality checks before delivering

- [ ] Governing thought on slide 1, not "Overview"
- [ ] Action titles (unless profile overrides)
- [ ] Every slide traces to key line
- [ ] Trust spine applied to exhibit claims

## Propose profile update

When a stable convention surfaces during this run (thresholds, naming, tone, output format, or recurring corrections), **propose a profile update**: show the exact diff against `~/.claude/plugins/config/claude-for-strategy/consulting/CLAUDE.md` (org-wide facts go to `org-profile.md`), ask for confirmation, and write only on yes. Only `/consulting:practice-setup` auto-applies a full profile write.

## Outputs

Follows plugin `CLAUDE.md` § Outputs. Next: approve outline → build slides, or `/consulting:so-what-sharpener` on weak titles.