tiktok-research（抖音研究）

TikTok Research

Research high-performing TikTok videos, identify outliers, and analyze top video content for hooks and structure.

Prerequisites

• APIFY_TOKEN environment variable or in .env
• GEMINI_API_KEY environment variable or in .env
• apify-client and google-genai Python packages
• Accounts configured in .claude/context/tiktok-accounts.md

Verify setup:

python3 -c "
import os
try:
    from dotenv import load_dotenv
    load_dotenv()
except ImportError:
    pass
from apify_client import ApifyClient
from google import genai
assert os.environ.get('APIFY_TOKEN'), 'APIFY_TOKEN not set'
assert os.environ.get('GEMINI_API_KEY'), 'GEMINI_API_KEY not set'
" && echo "Prerequisites OK"

Workflow

1. Create Run Folder

RUN_FOLDER="tiktok-research/$(date +%Y-%m-%d_%H%M%S)" && mkdir -p "$RUN_FOLDER" && echo "$RUN_FOLDER"

2. Fetch Content

python3 .claude/skills/tiktok-research/scripts/fetch_tiktok.py \
  --days 30 \
  --limit 50 \
  --sorting latest \
  --output {RUN_FOLDER}/raw.json

Parameters:

• --days: Days back to search (default: 30)
• --limit: Max videos per account (default: 50)
• --sorting: "latest", "popular", or "oldest" (default: latest)
• --usernames: Override accounts file with specific usernames

3. Identify Outliers

python3 .claude/skills/tiktok-research/scripts/analyze_posts.py \
  --input {RUN_FOLDER}/raw.json \
  --output {RUN_FOLDER}/outliers.json \
  --threshold 2.0

Output JSON contains:

• total_videos: Number of videos analyzed
• outlier_count: Number of outliers found
• topics: Top hashtags, sounds, and keywords
• accounts: List of accounts analyzed
• outliers: Array of outlier videos with engagement metrics

4. Analyze Top Videos with AI

python3 .claude/skills/video-content-analyzer/scripts/analyze_videos.py \
  --input {RUN_FOLDER}/outliers.json \
  --output {RUN_FOLDER}/video-analysis.json \
  --platform tiktok \
  --max-videos 5

Extracts from each video:

• Hook technique and replicable formula
• Content structure and sections
• Retention techniques
• CTA strategy

See the video-content-analyzer skill for full output schema and hook/format types.

5. Generate Report

Read {RUN_FOLDER}/outliers.json and {RUN_FOLDER}/video-analysis.json, then generate {RUN_FOLDER}/report.md.

Report Structure:

# TikTok Research Report

Generated: {date}

## Top Performing Hooks

Ranked by engagement. Use these formulas for your content.

### Hook 1: {technique} - @{username}
- **Opening**: "{opening_line}"
- **Why it works**: {attention_grab}
- **Replicable Formula**: {replicable_formula}
- **Engagement**: {diggCount} likes, {commentCount} comments, {playCount} views
- [Watch Video]({webVideoUrl})

[Repeat for each analyzed video]

## Content Structure Patterns

| Video | Format | Pacing | Key Retention Techniques |
|-------|--------|--------|--------------------------|
| @username | {format} | {pacing} | {techniques} |

## CTA Strategies

| Video | CTA Type | CTA Text | Placement |
|-------|----------|----------|-----------|
| @username | {type} | "{cta_text}" | {placement} |

## All Outliers

| Rank | Username | Likes | Comments | Shares | Views | Engagement Rate |
|------|----------|-------|----------|--------|-------|-----------------|
[List all outliers with metrics and links]

## Trending Topics

### Top Hashtags
[From outliers.json topics.hashtags]

### Top Sounds
[From outliers.json topics.sounds]

### Top Keywords
[From outliers.json topics.keywords]

## Actionable Takeaways

[Synthesize patterns into 4-6 specific recommendations]

## Accounts Analyzed
[List accounts]

Focus on actionable insights. The "Top Performing Hooks" section with replicable formulas should be prominent.

Quick Reference

Full pipeline:

RUN_FOLDER="tiktok-research/$(date +%Y-%m-%d_%H%M%S)" && mkdir -p "$RUN_FOLDER" && \
python3 .claude/skills/tiktok-research/scripts/fetch_tiktok.py -o "$RUN_FOLDER/raw.json" && \
python3 .claude/skills/tiktok-research/scripts/analyze_posts.py -i "$RUN_FOLDER/raw.json" -o "$RUN_FOLDER/outliers.json" && \
python3 .claude/skills/video-content-analyzer/scripts/analyze_videos.py -i "$RUN_FOLDER/outliers.json" -o "$RUN_FOLDER/video-analysis.json" -p tiktok

Then read both JSON files and generate the report.

Engagement Metrics

Engagement Score: likes + (3 x comments) + (2 x shares) + (2 x saves) + (0.05 x views)

Outlier Detection: Videos with engagement rate > mean + (threshold x std_dev)

Engagement Rate: (score / followers) x 100

TikTok-Specific Fields

• diggCount: Likes/hearts
• shareCount: Shares
• playCount: Video views
• commentCount: Comments
• collectCount: Saves/bookmarks
• authorFollowers: Creator's follower count
• musicName: Sound used in video
• musicOriginal: Whether sound is original