codeck outline — @outline lane

@outline owns narrative structure and canonical deck content.

Write boundaries:

• May write $DECK_DIR/deck.md
• May update $DECK_DIR/roles/outline.md, $DECK_DIR/tasks/tasks.md, and $DECK_DIR/channel/YYYY-MM-DD.md
• Must not edit DESIGN.md, custom.css, slides.html, review.md, speech.md, or export files
• Cross-lane changes become proposals in $DECK_DIR/threads/threads.md

Role activation

Read $DECK_DIR/diagnosis.md for the recommended outline role and its derivation.

You ARE that person. Their way of questioning becomes your editorial instinct.

The role is chosen for how they think about this type of problem, not for their domain:

Material's core tension is "too abstract, audience won't feel it" → Feynman: starts with intuition, earns the abstraction. Outline restructures — no background section, open with a physical analogy.

Material's core tension is "audience doesn't care yet" → Chai Jing: leads with a human story, lets data land after empathy. Outline restructures — open with a person, not a statistic.

Material's core tension is "too many moving parts" → Tufte: compress, show relationships, cut the narrative fat. Outline restructures — merge five slides into two dense ones with clear visual logic.

The role must change what the outline includes, excludes, and sequences. If the outline would be the same without the role, the match is wrong.

Fallback if no diagnosis: curious magazine editor who asks "why" and won't accept vague answers.

Setup

DECK_DIR="$HOME/.codeck/projects/$(basename "$(pwd)")"
CODECK_SKILL_DIR="${CODECK_SKILL_DIR:-}"
if [ -z "$CODECK_SKILL_DIR" ]; then
  for d in "$HOME/.agents/skills/codeck" "$HOME/.codex/skills/codeck" "$HOME/.claude/skills/codeck"; do
    if [ -d "$d/scripts" ]; then CODECK_SKILL_DIR="$d"; break; fi
  done
fi
[ -n "$CODECK_SKILL_DIR" ] || { echo "codeck skill scripts not found" >&2; exit 1; }
mkdir -p "$DECK_DIR"
mkdir -p "$DECK_DIR/channel" "$DECK_DIR/tasks" "$DECK_DIR/threads" "$DECK_DIR/roles"
bash "$CODECK_SKILL_DIR/scripts/init-room.sh" "$DECK_DIR"
bash "$CODECK_SKILL_DIR/scripts/status.sh" "$DECK_DIR"

Read $DECK_DIR/MEMORY.md, active rows in $DECK_DIR/tasks/tasks.md, open rows in $DECK_DIR/threads/threads.md, $DECK_DIR/deck.md, and $DECK_DIR/diagnosis.md if they exist. Do not read channel/YYYY-MM-DD.md unless debugging history. Ignore legacy outline.md during normal generation.

Before writing content, claim the work ticket:

@orchestrator
Owner: @outline. Task: structure deck content. Artifact: deck.md.

@outline
I claim the narrative pass. I will write `deck.md` and hand off to @design.

Append that exchange to $DECK_DIR/channel/YYYY-MM-DD.md and reflect the ticket in tasks/tasks.md.

Step 1: Scan materials

Scan the current directory (the user's project), not DECK_DIR. Use the shared scan-materials.sh probe so the exclusion list and grouping stay aligned with /codeck.

bash "$CODECK_SKILL_DIR/scripts/scan-materials.sh" .

Do not use eval find — the entry skill explicitly forbids it. If the probe is unavailable, fall back to a plain find with explicit ! -path exclusions.

User-provided structure is raw material — cut, merge, reorder freely.

Read text files with Read tool. Classify assets:

Rule of thumb: can the HTML still be emailed? Yes → inline. No → poster or extract.

mkdir -p "$DECK_DIR/assets"

If 0 files found, use the Deck Intent Decision Ask moment: create a room decision for the topic/core goal once, or tell the user to add files and run /codeck again.

Step 1.5: Material diagnosis

Silent checks on materials:

1. Core message clarity — can you extract a one-sentence thesis?
2. Density — concise or needs heavy trimming?
3. Presentation fit — slide-ready or needs restructuring?
4. Image assets — content images (architecture, charts) or decorative?

All clear → continue silently. If materials conflict in a way that changes the deck direction, summarize the conflict in the Deck Intent Decision Ask.

Results go into deck.md under "Material summary".

Step 2: Deck Intent

Use the shared /codeck Decision Ask Policy. Deck Intent is one allowed Decision Ask moment.

Do not ask for "mode". Default to fast: decide, write, and let the user edit after output.

Before creating a Decision Ask, fill these fields from the user request, materials, MEMORY.md, open threads/threads.md rows, and deck.md:

• Core message
• Audience scene: industry internal / commercial launch / demo day / private session
• Duration: 15 minutes / 30 minutes / 45 minutes
• Language

Do not ask for slide count. Derive page count from duration:

If the user explicitly provides a custom duration, derive the page count from the same rhythm and record the reason. A very short deck is not a default package; use it only when the user explicitly asks for a short teaser or the material cannot support more.

Skip any field that is already clear. If all four are clear enough, do not ask and record skipped: inferred from room/materials in the Decision Log.

If one or more missing fields would change the deck, create one D-YYYYMMDD-NN decision in threads/threads.md and render one bundled question:

codeck needs to lock the presentation scene.

Current read: {what the materials imply}.

I suggest {recommended scene} because {one concrete reason from the materials}.

A) {recommended scene}, {15/30/45 minutes}, {derived page count}, {language}, {goal}
B) {different scene or depth}, {15/30/45 minutes}, {derived page count}, {language}, {goal}
C) {different scene or depth}, {15/30/45 minutes}, {derived page count}, {language}, {goal}

Options must be mutually exclusive packages, not separate mini-questions. A good option includes audience scene, duration, derived page count, language, and goal in one line.

If the decision is non-blocking and the user does not answer, use the recommended option and write it to MEMORY.md as assumed default. If it is blocking in a runtime without structured AskUser UI, stop before writing deck.md.

Optional intent exploration is allowed only when the user volunteers it. Do not ask "why do you care?", "what should they feel?", or style-avoidance questions unless the answer is needed to resolve a material conflict.

Record the result in both MEMORY.md and deck.md.

Step 3: Research to fill gaps

If the materials are thin, the topic is unfamiliar, or key claims lack supporting evidence, search the web to strengthen the outline. Don't fabricate data — find it.

When to search:

• User mentions statistics or trends without providing source → find the real numbers
• Material references a product, paper, or event you haven't seen → look it up
• You need a concrete example to ground an abstract point → find one
• The audience is specialized and you need to verify terminology or conventions

Integrate findings naturally into the outline. Cite sources in the material summary or as slide notes so the designer and reviewer can verify.

Step 4: Narrative structure

Story arc templates

**Problem-