Phase 7: Post-Event — Detail

Triggered separately — days or weeks after delivery. Not part of the linear Phase 0-6 flow. The talk has been given, the recording exists, and the speaker wants a YouTube thumbnail and/or to update shownotes with the video.

Pre-Flight Checklist

Before ANY Phase 7 action, load these files. If any is missing, STOP and ask.

1. speaker-profile.json — thumbnail preferences, video publishing config
2. secrets.json — API keys (Gemini for thumbnail generation)
3. presentation-spec.md — talk slug, metadata (source of truth)
4. presentation-outline.md — the outline (slide references, illustration references)
5. YouTube video URL — provided by the speaker at trigger time

If shownotes don't exist and the speaker wants Step 7.2, STOP and ask — either run Phase 6 Step 6.1 first, or get the shownotes URL manually.

Step 7.1: YouTube Thumbnail Generation

Generate a thumbnail that competes for clicks in YouTube search results and suggested videos. The thumbnail combines a slide visual, the speaker's face, and a short hook title.

1. Slide Selection

Scan the outline for high-impact slides. Suggest 3-5 candidates ranked by visual engagement:

• Illustrations — highest impact, already designed for visual punch
• Bold claims / provocative statements — text that triggers curiosity
• Key diagrams / architecture visuals — concrete and recognizable
• Demo screenshots — when the demo is the main attraction

Avoid: bio slides, shownotes URL slides, bullet-heavy slides, generic titles.

Present candidates to the speaker with slide numbers and brief descriptions. The speaker picks one — never auto-select.

2. Slide Image Resolution

Get the slide image using this resolution chain:

1. Existing illustration (preferred) — check illustrations/ directory for slide-{NN}.* matching the chosen slide. These are already high-quality.
2. PPTX extraction — use the helper mode:

   python3 skills/presentation-creator/scripts/generate-thumbnail.py \
     --extract-slide deck.pptx 15 --output slide-15.png

   Uses LibreOffice headless or PowerPoint AppleScript on macOS.
3. Ask user — if extraction fails, ask the speaker for a screenshot or exported image of the slide.

3. Speaker Photo

Resolution order:

1. publishing_process.thumbnail.speaker_photo_path from speaker profile
2. Ask the speaker to provide a path or URL

The photo must be a real photograph — never AI-generated. Expression should convey engagement and energy, not a neutral corporate headshot.

4. Thumbnail Title Text

This is NOT the full talk title. It's a 3-5 word hook designed for thumbnail readability at small sizes.

• Propose 2-3 options based on the talk's thesis and key claim
• The speaker confirms or edits the text
• ALL CAPS is standard for YouTube thumbnails
• Must be readable when the thumbnail is displayed at 160x90 pixels

5. Generate

Decide the recommended --aesthetic per the precedence chain in rules/thumbnail-generation-rules.md Rule 7. Walk these in order; the first match wins:

1. publishing_process.thumbnail.aesthetic_preference — explicit speaker preference ("photo" or "comic_book"). Honor it and stop.
2. visual_style_history.default_illustration_style — observed pattern from past talks. Fuzzy-match: matches the comic-book family (comic_book / comic-book / halftone / illustrated / cartoon / caricature) → recommend comic_book. Matches a different documented style (retro_tech_manual, watercolor, etc.) → out-of-scope for current aesthetics; ask the speaker before generating. Otherwise fall through.
3. visual_style_history.confirmed_visual_intents — same fuzzy-match against each entry's pattern and rule fields.
4. Default → photo.

Lead with the recommended aesthetic as the primary candidate. Offer a two-candidate side-by-side comparison only when the speaker is genuinely undecided or wants to validate before committing — not as a default. The comic-book treatment is high-variance: when it works it produces significantly higher CTR than photo composites, when it misses it looks off-brand.

Pass through the deck's Illustration Style Anchor. If presentation-outline.md contains a ## STYLE ANCHOR block (Phase 2's illustration strategy output), pass its content through to generate-thumbnail.py via --portrait-style "<anchor>". The script will pre-stylize the speaker photo into the anchor's medium (sepia tech-manual, watercolor, pen-and-ink, etc.) before composition, so the output palette matches the deck instead of clashing with it. Without this pass-through, photographic skin tones beside an illustrated background produce a jarring two-medium composite even when the selected aesthetic is otherwise correct. If Phase 2 didn't produce a style anchor (e.g., stock-image-only deck), omit --portrait-style.

# Option A: photographic composite (conservative)
python3 skills/presentation-creator/scripts/generate-thumbnail.py \
  --slide-image illustrations/slide-15.png \
  --speaker-photo ~/photos/headshot.jpg \
  --title "JUDGMENT DAY" \
  --subtitle "DevNexus 2026" \
  --vault ~/.claude/rhetoric-knowledge-vault \
  --aesthetic photo \
  --style slide_dominant \
  --title-position top \
  --brand-colors "#5B2C6F,#C0392B" \
  --output thumbnail-photo.png

# Option B: comic-book caricature (viral)
python3 skills/presentation-creator/scripts/generate-thumbnail.py \
  --slide-image illustrations/slide-15.png \
  --speaker-photo ~/photos/headshot.jpg \
  --title "JUDGMENT DAY" \
  --subtitle "DevNexus 2026" \
  --vault ~/.claude/rhetoric-knowledge-vault \
  --aesthetic comic_book \
  --style slide_dominant \
  --title-position top \
  --brand-colors "#5B2C6F,#C0392B" \
  --output thumbnail-comic.png

# Option C: anchor-matched (when Phase 2 produced a STYLE ANCHOR)
python3 skills/presentation-creator/scripts/generate-thumbnail.py \
  --slide-image illustrations/slide-15.png \
  --speaker-photo ~/photos/headshot.jpg \
  --title "JUDGMENT DAY" \
  --subtitle "DevNexus 2026" \
  --vault ~/.claude/rhetoric-knowledge-vault \
  --aesthetic photo \
  --portrait-style "Retro U.S. Military WWII technical manual, sepia, pen-and-ink crosshatching, WWII uniform with garrison cap" \
  --style slide_dominant \
  --title-position top \
  --output thumbnail-anchored.png

For most runs only ONE candidate is needed — the one chosen by the precedence chain above. The two-candidate command set is only for genuine uncertainty.

Apply other speaker preferences from publishing_process.thumbnail:

• style_preference → --style
• title_position → --title-position
• brand_colors → --brand-colors

(aesthetic_preference is consumed at the top of this step as the first entry in the precedence chain — don't re-apply it here.)

The script (per invocation — runs once per --aesthetic):

• Sends the slide image + speaker photo + prompt to Gemini as multimodal input (two input images, one Gemini call)
• Uses researched prompt strategy per the chosen aesthetic
• Validates output: exactly 1280x720, <2MB, PNG preferred
• Saves to the specified output path (default: thumbnail.png in illustrations dir)

6. Speaker Review

Present the generated thumbnail for approval. If rejected:

• Adjust specific prompt components (style, position, colors, text)
• Try a different slide image
• Do NOT regenerate from scratch — iterate on the existing prompt

Common revision requests:

• "Face looks wrong" → strengthen face preservation language
• "Text is unreadable" → increase contrast, change position
• "Too busy" → switch to simpler style variant
• "Wrong mood" → adjust expression guidance

7. Copy Thumbnail to Shownotes Site

If publishing_process.shownotes.enabled is true, the SSG template expects the thumbnail at a specific path relative to the shownotes site root — otherwise the live page falls back to a placeholder image with no warning.

Resolve the destination using publishing_process.shownotes:

{shownotes.source.path_or_url}/{shownotes.thumbnail_path_template}

with {slug} substituted from the Presentation Spec. For Jekyll-based shownotes in this toolkit the default template is assets/images/thumbnails/{slug}-thumbnail.png — both the nested thumbnails/ subdirectory AND the -thumbnail suffix are mandatory. Do not strip either when landing the file.

Create the thumbnails/ directory if it doesn't exist, then copy (don't move) the generated thumbnail — the local copy in illustrations/thumbnail.png stays with the talk's working directory for tracking.

If the SSG template pointer (shownotes.ssg_template_pointer) is set, read that file after a site redesign to re-derive the path convention. Don't re-invent it from folklore.

8. Tracking Database Update

Add to thumbnails[] in tracking-database.json:

{
  "talk_slug": "judgment-day",
  "youtube_url": "https://youtube.com/watch?v=...",
  "source_slide_num": 15,
  "speaker_photo_used": "/path/to/headshot.jpg",
  "thumbnail_path": "illustrations/thumbnail.png",
  "shownotes_thumbnail_path": "assets/images/thumbnails/judgment-day-thumbnail.png",
  "dimensions": "1280x720",
  "file_size_kb": 185,
  "created_at": "2026-04-20",
  "approved": true
}

Also set thumbnail_generated: true on the talk's entry in talks[].

Step 7.2: Video to Shownotes

Update the existing shownotes page with the video recording link.

1. Verify Shownotes Exist

Check that shownotes were published in Phase 6 Step 6.1. Look for:

• The shownotes URL in tracking-database.json for this talk
• Or construct from publishing_process.shownotes.url.base + publishing_process.shownotes.url.template (substitute {slug} and any date variables) — see phase6-publishing.md for the full template semantics

If shownotes don't exist, STOP and ask the speaker. Options:

• Run Phase 6 Step 6.1 to create them now
• Provide the shownotes URL manually
• Skip this step

2. Read Video Publishing Config

Read publishing_process.video_publishing from speaker profile:

• enabled — if false, skip this step
• embed_method — youtube_embed, link_only, or both
• shownotes_video_section — where/how the video section goes in shownotes
• video_description_template — template for YouTube video description

If video_publishing is not configured, ask interactively:

• "How should the video appear in shownotes? (embed, link, or both)"
• "Where in the shownotes should it go?"

3. Generate Shownotes Update

Create the video section content based on embed_method:

• youtube_embed: Full responsive YouTube embed iframe
• link_only: "Watch the Recording" heading with a link to the video
• both: Embed iframe + text link below

Include the video title, conference name, and year from the presentation spec.

4. Apply Update

Use the same publishing method as Phase 6 Step 6.1:

• If git-based: create/update the shownotes file, commit, push
• If CMS: provide the content for the speaker to paste
• If manual: present the formatted content block

5. Tracking Database Update

Set video_added_to_shownotes: true on the talk's entry in talks[].

Add the YouTube URL to the talk entry if not already present.

Phase 7 Report

POST-EVENT REPORT — {talk title}
==================================
[DONE/SKIP] Thumbnail: {path, dimensions, size}
[DONE/SKIP] Video to shownotes: {shownotes URL, embed method}
[INFO] YouTube URL: {url}
==================================