Self-evolving trend prediction for content creators. Monitor 11 platforms. Get told exactly what to post and when. Watch the model improve every week.
[Full technical documentation → SKILL.md]
# 1. Install
clawhub install 0xF69/viralevo
# 2. Add your Tavily API key (free at tavily.com)
echo "TAVILY_API_KEY=tvly-xxxx" >> ~/.openclaw/workspace/.env
# 3. Run setup check
python3 ~/.openclaw/workspace/viralevo/setup.py
# 4. Start onboarding — tell your agent:
# "Start ViralEvo setup"
# — or in Chinese —
# "开始趋势雷达设置"| ViralEvo | Generic Trend Tools | Manual Scrolling | |
|---|---|---|---|
| Catches trends before mainstream | ✅ 12–48h early | ❌ Lags | ❌ |
| Learns from your specific results | ✅ | ❌ | ❌ |
| Self-corrects weekly | ✅ | ❌ | ❌ |
| Transparent reasoning | ✅ Full evidence trail | ❌ Black box | ✅ |
| Chinese language support | ✅ | ❌ | ✅ |
| Cost | Free (Tavily free tier) | $29–$299/mo | Your time |
AI/Tech · E-commerce · Beauty · Fitness · Finance · Gaming · Fashion · Education · Real Estate · Pets · Custom
HackerNews · Dev.to · Reddit · Product Hunt · YouTube · Twitter/X · TikTok · Instagram · Pinterest · LinkedIn
| Period | Expected Accuracy |
|---|---|
| Week 1–2 | 30–40% (cold start) |
| Month 2 | 55–65% |
| Month 3+ | 65–75% |
| Month 6+ | 75%+ |
Accuracy = prediction within ±20% of actual topic lifecycle.
All collected data and prediction models are stored locally on your machine. No user data is ever transmitted externally. The skill makes outbound requests only to fetch public trend signals (HackerNews, Dev.to, Reddit, Product Hunt, Tavily). Tavily receives only search query strings.
- Node.js v18+
- Python 3.10+
- OpenClaw v2026.1+
- Tavily API key (free tier sufficient — get one at tavily.com)
MIT — see LICENSE
See SKILL.md for complete technical documentation, scoring formula, disclaimers, and sample reports.