Generate Output
Generate one of 8 output types from ingested document sources.
Output types
| Type |
Command |
Output files |
flashcards |
/notebook:generate flashcards [topic] |
.json + .md + .html (interactive flip cards) |
quiz |
/notebook:generate quiz [topic] |
.json + .md + .html (50 Qs, live scoring) |
report |
/notebook:generate report [topic] |
.md + .docx |
slide-deck |
/notebook:generate slide-deck [topic] |
.pptx |
mind-map |
/notebook:generate mind-map [topic] |
.mmd + .html (Mermaid viewer) |
infographic |
/notebook:generate infographic [topic] |
.html |
data-table |
/notebook:generate data-table [topic] |
.csv + .json + .md + .html (sortable/filterable) |
audio-overview |
/notebook:generate audio-overview [topic] |
.wav (Kokoro TTS) |
Workflow
- Retrieve context — query ingested sources:
bun "${CLAUDE_PLUGIN_ROOT}/skills/ingest/scripts/query.mjs" "$ARGUMENTS" --top-k 15
- Build input JSON — see the relevant template in assets/ for the schema:
| Type |
Template |
Input key fields |
| flashcards |
assets/flashcards.json |
[{ "front", "back", "tags" }] |
| quiz |
assets/quiz.json |
{ "title", "questions": [{ "question", "type", "options", "answer", "explanation" }] } |
| report |
assets/report-docx.json |
{ "title", "sections": [{ "heading", "content" }] } |
| slide-deck |
assets/slide-deck.json |
{ "title", "slides": [{ "title", "bullets"/"body", "notes" }] } |
| mind-map |
assets/mind-map.mmd |
{ "label", "children": [...] } or raw Mermaid |
| infographic |
assets/infographic.json |
{ "title", "sections": [{ "heading", "content", "icon", "stat" }] } |
| data-table |
assets/data-table.json |
{ "headers": [...], "rows": [[...]] } |
| audio-overview |
assets/audio-overview.json |
{ "segments": [{ "text", "voice" }] } |
Write input JSON to output/_<type>_input.json.
- Run the generator:
# Flashcards (zero-dep)
node "${CLAUDE_SKILL_DIR}/scripts/flashcards.mjs" -i <input> -o output --name flashcards
# Quiz (zero-dep)
node "${CLAUDE_SKILL_DIR}/scripts/quiz.mjs" -i <input> -o output --name quiz
# Report (bun auto-installs deps)
bun "${CLAUDE_SKILL_DIR}/scripts/report.mjs" -i <input> -o output --name report --format both
# Slide deck (bun auto-installs deps)
bun "${CLAUDE_SKILL_DIR}/scripts/slide-deck.mjs" -i <input> -o output --name slide-deck
# Mind map (zero-dep)
node "${CLAUDE_SKILL_DIR}/scripts/mind-map.mjs" -i <input> -o output --name mind-map
# Infographic (zero-dep)
node "${CLAUDE_SKILL_DIR}/scripts/infographic.mjs" -i <input> -o output --name infographic
# Data table (bun auto-installs deps)
bun "${CLAUDE_SKILL_DIR}/scripts/data-table.mjs" -i <input> -o output --name data-table
# Audio overview (Kokoro TTS neural voice via uv, falls back to macOS say)
uv run --script "${CLAUDE_SKILL_DIR}/scripts/audio-overview.py" -i <input> -o output --name audio-overview
- Report output paths. For HTML outputs, tell user:
open output/<name>.html
Quality gates
- Flashcards: 15-30 cards, no yes/no questions, tagged by topic
- Quiz: 50 questions, 70% MC / 30% SA, explanations for all
- Report: executive summary, evidence-backed findings, actionable conclusions
- Slide deck: 8-15 content slides, 3-5 bullets each, speaker notes
- Mind map: 3-6 main branches, 2-4 sub-topics each, concise labels
- Infographic: 5-8 sections, stats highlighted, concise
- Data table: specific data points, consistent column formatting
- Audio: 8-15 alternating segments, conversational tone, 2-4 sentences each
Gotchas
- Always query sources first — ground all content in source material
- Slide deck:
bullets and body are mutually exclusive per slide
- Audio: uses Kokoro TTS (82M neural model). Best voices:
af_heart (female, A-grade), am_fenrir (male, C+). Falls back to macOS say if Kokoro not installed. Strip URLs and special characters.
- Mind map: Mermaid syntax is indentation-sensitive, avoid
()[] in labels
- Report DOCX: supports paragraphs and headings only (no tables/images)
- Quiz MC:
answer field is the letter (A/B/C/D), not the full text
Additional resources
- Voice options for audio: references/voices.md
$ARGUMENTS
1---2name: generate3description: Generate rich outputs from ingested sources — flashcards, quizzes, reports, slide decks, mind maps, infographics, data tables, or audio overviews. Use when the user wants to create any output from their documents.4---56# Generate Output78Generate one of 8 output types from ingested document sources.910## Output types1112| Type | Command | Output files |13|------|---------|-------------|14| `flashcards` | `/notebook:generate flashcards [topic]` | .json + .md + .html (interactive flip cards) |15| `quiz` | `/notebook:generate quiz [topic]` | .json + .md + .html (50 Qs, live scoring) |16| `report` | `/notebook:generate report [topic]` | .md + .docx |17| `slide-deck` | `/notebook:generate slide-deck [topic]` | .pptx |18| `mind-map` | `/notebook:generate mind-map [topic]` | .mmd + .html (Mermaid viewer) |19| `infographic` | `/notebook:generate infographic [topic]` | .html |20| `data-table` | `/notebook:generate data-table [topic]` | .csv + .json + .md + .html (sortable/filterable) |21| `audio-overview` | `/notebook:generate audio-overview [topic]` | .wav (Kokoro TTS) |2223## Workflow24251. **Retrieve context** — query ingested sources:2627```bash28bun "${CLAUDE_PLUGIN_ROOT}/skills/ingest/scripts/query.mjs" "$ARGUMENTS" --top-k 1529```30312. **Build input JSON** — see the relevant template in [assets/](assets/) for the schema:3233| Type | Template | Input key fields |34|------|----------|-----------------|35| flashcards | [assets/flashcards.json](assets/flashcards.json) | `[{ "front", "back", "tags" }]` |36| quiz | [assets/quiz.json](assets/quiz.json) | `{ "title", "questions": [{ "question", "type", "options", "answer", "explanation" }] }` |37| report | [assets/report-docx.json](assets/report-docx.json) | `{ "title", "sections": [{ "heading", "content" }] }` |38| slide-deck | [assets/slide-deck.json](assets/slide-deck.json) | `{ "title", "slides": [{ "title", "bullets"/"body", "notes" }] }` |39| mind-map | [assets/mind-map.mmd](assets/mind-map.mmd) | `{ "label", "children": [...] }` or raw Mermaid |40| infographic | [assets/infographic.json](assets/infographic.json) | `{ "title", "sections": [{ "heading", "content", "icon", "stat" }] }` |41| data-table | [assets/data-table.json](assets/data-table.json) | `{ "headers": [...], "rows": [[...]] }` |42| audio-overview | [assets/audio-overview.json](assets/audio-overview.json) | `{ "segments": [{ "text", "voice" }] }` |4344Write input JSON to `output/_<type>_input.json`.45463. **Run the generator**:4748```bash49# Flashcards (zero-dep)50node "${CLAUDE_SKILL_DIR}/scripts/flashcards.mjs" -i <input> -o output --name flashcards5152# Quiz (zero-dep)53node "${CLAUDE_SKILL_DIR}/scripts/quiz.mjs" -i <input> -o output --name quiz5455# Report (bun auto-installs deps)56bun "${CLAUDE_SKILL_DIR}/scripts/report.mjs" -i <input> -o output --name report --format both5758# Slide deck (bun auto-installs deps)59bun "${CLAUDE_SKILL_DIR}/scripts/slide-deck.mjs" -i <input> -o output --name slide-deck6061# Mind map (zero-dep)62node "${CLAUDE_SKILL_DIR}/scripts/mind-map.mjs" -i <input> -o output --name mind-map6364# Infographic (zero-dep)65node "${CLAUDE_SKILL_DIR}/scripts/infographic.mjs" -i <input> -o output --name infographic6667# Data table (bun auto-installs deps)68bun "${CLAUDE_SKILL_DIR}/scripts/data-table.mjs" -i <input> -o output --name data-table6970# Audio overview (Kokoro TTS neural voice via uv, falls back to macOS say)71uv run --script "${CLAUDE_SKILL_DIR}/scripts/audio-overview.py" -i <input> -o output --name audio-overview72```73744. Report output paths. For HTML outputs, tell user: `open output/<name>.html`7576## Quality gates7778- **Flashcards**: 15-30 cards, no yes/no questions, tagged by topic79- **Quiz**: 50 questions, 70% MC / 30% SA, explanations for all80- **Report**: executive summary, evidence-backed findings, actionable conclusions81- **Slide deck**: 8-15 content slides, 3-5 bullets each, speaker notes82- **Mind map**: 3-6 main branches, 2-4 sub-topics each, concise labels83- **Infographic**: 5-8 sections, stats highlighted, concise84- **Data table**: specific data points, consistent column formatting85- **Audio**: 8-15 alternating segments, conversational tone, 2-4 sentences each8687## Gotchas8889- Always query sources first — ground all content in source material90- Slide deck: `bullets` and `body` are mutually exclusive per slide91- Audio: uses Kokoro TTS (82M neural model). Best voices: `af_heart` (female, A-grade), `am_fenrir` (male, C+). Falls back to macOS `say` if Kokoro not installed. Strip URLs and special characters.92- Mind map: Mermaid syntax is indentation-sensitive, avoid `()[]` in labels93- Report DOCX: supports paragraphs and headings only (no tables/images)94- Quiz MC: `answer` field is the letter (A/B/C/D), not the full text9596## Additional resources9798- Voice options for audio: [references/voices.md](references/voices.md)99100$ARGUMENTS
damionrashford