viralevo_skill

Overview

This skill is a self-evolving viral content trend advisor that monitors 11 platforms, scores emerging topics, and recommends what to post and when. It continuously improves accuracy by validating predictions against your reported results and automatically adjusts weighting each week. Designed to run inside OpenClaw with local storage and optional Tavily integration.

How this skill works

ViralEvo collects signal data from multiple public sources and Tavily search, computes a weighted score for topics, and estimates lifecycle windows. It verifies predictions at 5 and 65 minutes after reports and aggregates weekly errors to adjust scoring weights automatically. Reports and health metrics are written locally; you can feed post outcomes back to refine model performance.

When to use it

• Daily content planning to decide what to publish today
• Monitoring cross-platform trend signals for early opportunities
• Automating weekly model tuning without manual weight updates
• Verifying prediction accuracy after posting and logging results
• Running scheduled collect → report via OpenClaw cron

Best practices

• Complete onboarding and set your preferred niche and language for tailored signals
• Add your Tavily API key to avoid search limits and improve coverage
• Report real post outcomes (views, saves, engagement) to close the feedback loop
• Schedule the four recommended cron jobs in OpenClaw for daily and weekly automation
• Run weekly_review manually if accuracy drops or after major platform changes

Example use cases

• A creator asks “What should I post today?” and receives ranked topics with posting windows
• A marketing team uses daily reports to align social posts across channels for a campaign
• A small publisher monitors niche signals to catch trends 12–48 hours earlier than competitors
• An ops user automates weekly weight tuning to maintain model accuracy over months