agentleFS
kuker24kuker24/OpenCodeBestFriend/skills/id-demo-video/SKILL.md

id-demo-video

Skill0 starsChanged 2 months agoMITView on GitHub

Use when the user wants an application demo video, product walkthrough recording, lomba/contest demo, Indonesian voiceover for a screen demo, or a ~10 minute narrated app tour. Orchestrates beat sheet, Indonesian narration, Edge TTS, Playwright/Chromium scene capture, ffmpeg concat, optional HyperFrames title cards. Not for functional browser QA (playwright-qa), UI motion inside the product (emil-design-eng), still/ad art (visual-studio), file+line developer tours (code-tour), or a single HTML composition standing in for a 10 minute video (hyperframes handles title/end cards only).

  • Installs packages

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 5 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    App Availability Gate: If the target application cannot be contacted, fail immediately and instruct the user to run the app start command.

    line 28
  2. Distinctive:

    No Secrets on Screen: Only seed and demo data may be typed or displayed.

    line 30
  3. Distinctive:

    Natural Spoken Indonesian: Write conversational Indonesian with short sentences (8–16 words).

    line 36
  4. Distinctive:

    Anti-Slop Buzzword Prohibition: Do not use empty corporate jargon:

    line 37
  5. Very specific:

    All demo assets reside in the project's demos/<slug>/ directory (must be gitignored):

    line 43
  6. Very specific:

    Naskah Indonesia: Write natural spoken Indonesian voiceover per scene in naskah.md (130–150 wpm, 8–16 words/sentence).

    line 64
  7. Very specific:

    Edge TTS Synthesis: Run python3 scripts/tts_edge.py to synthesize audio/scene-NN.mp3 with content-hashing.

    line 66
  8. Very specific:

    Compose & Concat: Run scripts/compose.sh to trim idle gaps (>1.2s), mux audio, burn subtitles, and concatenate scenes into demo-final.mp4.

    line 69
  9. Distinctive:

    Verify environment tools before running scripts:

    line 74

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Indonesian Application Demo Video (id-demo-video)

Director and orchestrator for multi-scene application walkthrough videos with natural Indonesian narration, screen recording, synchronized subtitles, and zero-cost local tooling.

Boundaries & Handoffs

Request / RequirementPrimary Specialist
Functional browser QA, regression assertions, testing form flowsplaywright-qa
In-app UI micro-interactions, CSS transitions, easing curvesemil-design-eng (after impeccable)
Photoreal product stills, marketing ad creative packs, non-UI surfacesvisual-studio
Developer onboarding and interactive CodeTour .tour file walkthroughscode-tour
Deterministic HTML title cards, lower-thirds, or short (<90s) motion cardshyperframes (handoff for intro/outro cards)
Full application demo video, screen walkthrough, and Indonesian narrationid-demo-video

Core Architecture & Fences

  1. Pipeline Director: id-demo-video coordinates the demo creation pipeline. It is not an image generator or a single-composition renderer.
  2. HyperFrames Boundary: HyperFrames handles short HTML-to-MP4 title/end cards (sweet spot 30–90s). Never attempt a single 600s monolithic HTML composition for a 10-minute demo.
  3. Free & Local Default (Rp0): Edge TTS with voices id-ID-GadisNeural (female) or id-ID-ArdiNeural (male). No paid API keys required. Kokoro is prohibited for Indonesian text. ElevenLabs is strictly OFF by default (if explicitly turned on by user, warn about ~10k free credit limits and non-commercial restrictions).
  4. App Availability Gate: If the target application cannot be contacted, fail immediately and instruct the user to run the app start command. Never render a fake dummy screen or blank placeholder video.
  5. Session & Browser Rules: Record by launching a dedicated, standalone Playwright Chromium process with an isolated profile. Never bind to personal Google Chrome profiles. Do not connect to or interfere with port 9223 (reserved for chrome-devtools-axi QA sessions; recording must run its own independent browser process).
  6. No Secrets on Screen: Only seed and demo data may be typed or displayed. Mask or clear all private keys, passwords, and tokens.

Ten-Minute Demo Rules

  • Modular Scene Architecture: Split a ~10-minute (600s) target into 8–12 discrete scenes (40–90 seconds each). One continuous 10-minute take is strictly prohibited.
  • Pacing & Cadence: Speaking rate is 130–150 words per minute. A 10-minute video contains ~1,300–1,500 spoken words.
  • Natural Spoken Indonesian: Write conversational Indonesian with short sentences (8–16 words).
  • Anti-Slop Buzzword Prohibition: Do not use empty corporate jargon:
    • Banned terms: "era digital", "solusi inovatif", "memanfaatkan AI", "seamless", "cutting-edge", "game changer".
    • Focus on what the user does, what appears on screen, and why it matters.

Artifact Contract

All demo assets reside in the project's demos/<slug>/ directory (must be gitignored):

demos/<slug>/
  brief.md                  # Scope, URL, feature checklist, target duration
  beats.md                  # Timed scene outline (total seconds = target ±15s)
  naskah.md                 # Per-scene spoken Indonesian voiceover script
  storyboard.json           # Machine-executable scene actions & audio mapping
  audio/scene-NN.mp3        # Synthesized voiceover per scene (content-cached)
  recordings/scene-NN.webm  # Raw Playwright screen capture per scene
  captions/scene-NN.srt     # Synchronized Indonesian subtitle per scene
  scenes/scene-NN.mp4       # Composed per-scene video (audio + video + subs)
  demo-preview-90s.mp4      # 90-second executive cut (intro + core + outro)
  demo-final.mp4            # Full concatenated walkthrough
  QC.md                     # Quality checklist and duration report

Nine-Step Execution Pipeline

  1. Brief: Extract app purpose, target audience, local URL (e.g. http://localhost:3000), and mandatory features from repo files or user prompt into brief.md.
  2. Beat Sheet: Design timed scene beats in beats.md. Ensure sum(scene_durations) == target_duration ± 15s. Consult references/beats.md.
  3. Naskah Indonesia: Write natural spoken Indonesian voiceover per scene in naskah.md (130–150 wpm, 8–16 words/sentence).
  4. Storyboard: Structure actions into storyboard.json. Consult references/storyboard.schema.md.
  5. Edge TTS Synthesis: Run python3 scripts/tts_edge.py to synthesize audio/scene-NN.mp3 with content-hashing. Measure exact audio duration with ffprobe before recording. Consult references/tts-id.md.
  6. Scene Recording: Drive browser via Playwright CLI with 1920x1080 viewport, visible cursor, and human typing cadence. Capture raw video into recordings/scene-NN.webm. Consult references/record-compose.md.
  7. Subtitle Generation: Generate synchronized Indonesian subtitles (captions/scene-NN.srt) matching speech segments.
  8. Compose & Concat: Run scripts/compose.sh to trim idle gaps (>1.2s), mux audio, burn subtitles, and concatenate scenes into demo-final.mp4.
  9. Preview & QC: Generate demo-preview-90s.mp4 containing representative scenes and record verification results in QC.md.

Runtime Prerequisites

Verify environment tools before running scripts:

ffmpeg -version >/dev/null 2>&1 || { echo "MISSING: ffmpeg"; exit 1; }
ffprobe -version >/dev/null 2>&1 || { echo "MISSING: ffprobe"; exit 1; }
python3 -c "import edge_tts" >/dev/null 2>&1 || { echo "MISSING: edge-tts (install via: pip install edge-tts)"; exit 1; }

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.