tiktok-ad-analysis
AUTOMATIC video market analysis with hybrid transcription (TikTok captions → Whisper fallback). Auto-triggers when videos exist in ref_video/. Analyzes TikTok reference videos to extract hooks, strategies, and insights for script generation. OPTIMIZED for speed with parallel processing across products.
npx skills add majiayu000/claude-skill-registry --skill tiktok-ad-analysis --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
What it does
Automatically analyzes TikTok reference videos with bilingual output (English + Chinese)
How it works
- Triggers automatically after Step 1 scraping completes if ref_video contains .mp4 files.
- For each video in ref_video, the Python script extracts frames (keyframes) and audio, attempts to fetch captions via yt-dlp, falls back to Whisper transcription if needed, then passes frames + transcript to Gemini for analysis.
- Outputs per-video analysis files named video_N_analysis.md and a mandatory market summary file video_synthesis.md.
When to use it
Auto-triggers when there are videos in ref_video and the base path exists; manual invocation is described in the script usage for single or parallel batch processing.
What it can touch
- Files under product_list/YYYYMMDD/{product_id}/ref_video/: video_N_analysis.md, video_synthesis.md, temp folders for frames and audio during processing.
- Scripts: analyze_video_batch.py, analyze_single_video.py perform the analysis workflow.
- Tools: FFmpeg, yt-dlp, Whisper (faster-whisper), Gemini via gemini-cli, Python environment with specified dependencies.
Caveats
- Output is bilingual: English + Chinese and includes a per-video analysis plus a market synthesis; a Compliance & Trust Signals section is required when writing video_synthesis.md.
- Transcription may fall back from TikTok captions to Whisper; if both fail, the transcript is marked unavailable.
- Parallelization is limited to up to 5 products concurrently as per the integration notes.
- Requires ref_video existence and that tabcut_data.json is available for performance metadata.
# TikTok Ad Analysis Skill (AUTOMATIC) **WHAT THIS DOES:** Automatically analyzes TikTok reference videos with bilingual output (English + Chinese) **WHEN IT RUNS:** AUTO-TRIGGERS after Step 1 (scraping) if `ref_video/` folder contains .mp4 files **HOW IT WORKS:** Python script extracts frames + transcribes audio → Gemini generates analysis **OUTPUT:** `video_N_analysis.md` for each video + `video_synthesis.md` summary (MANDATORY) --- ## Compliance & Policy Notes (DE Market) This is an *analysis* skill, but it should proactively flag compliance-risk claims found in source videos so the script generator can avoid them. - **Price / discount claims:** flag exact `€` amounts / “Euro” / “欧元”, hard discounts, and precise comparisons. - **Waterproof claims:** flag absolutes like `"100% wasserdicht"`, `"komplett wasserdicht"`, `"完全防水"` unless an IP rating is visible in packaging. - **Medical claims:** flag pain/therapy/healing language (e.g., `Schmerzfreiheit`, `Therapeut`, `Physio`, `heilt`). - **Tech specs:** flag ambiguous claims like `"4K Support"` (often decode, not native). When writing `video_synthesis.md`, include a small “Compliance & Trust Signals” section listing what to avoid i
- Compliance & Policy Notes (DE Market)
- Integration with Skill Pipeline
- Auto-Trigger Logic
- Prerequisites
- Transcription Workflow (Hybrid Approach)
- Script Files
- analyzevideobatch.py - Batch Processing
- analyzesinglevideo.py - Single Video
- Analysis Output Format
- 1. Video Metadata | 视频元数据
- 2. Voiceover/Dialogue Transcript | 旁白/对话文本
- Chinese Translation Philosophy | 中文翻译哲学
- 3. Hook/Opening Strategy | 开场策略
- 4. Shot-by-Shot Storyboard | 分镜脚本
After Step 1 (product scraping) completes:
if [ -d "$base/ref_video" ]; then
if [ $video_count -gt 0 ]; then
echo "✅ AUTO-TRIGGER: Found $video_count videos"
python analyze_video_batch.py {product_id} --date "$date"
else
echo "⏭️ SKIP: No videos found"
fi
echo "⏭️ SKIP: No ref_video folder"
Navigate to scripts directoryWhat does the tiktok-ad-analysis skill do?
AUTOMATIC video market analysis with hybrid transcription (TikTok captions → Whisper fallback). Auto-triggers when videos exist in ref_video/. Analyzes TikTok reference videos to extract hooks, strategies, and insights for script generation. OPTIMIZED for speed with parallel processing across products.
How do I install it?
Run `npx skills add majiayu000/claude-skill-registry --skill tiktok-ad-analysis --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From majiayu000/claude-skill-registry, a repository with 534 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.