An AI agent skill that queries multiple disconnected platforms in parallel, ranks results using human engagement signals, and synthesizes a brief. It requires external APIs and internet access to operate.
Project overview
Reduces manual, fragmented research by clustering stories across platforms and scoring them using live engagement signals before an AI judge synthesizes the findings.
Project type
AI Coding
Deployment
Refer to project documentation
License
MIT
Best for
Developers using CLI-based AI agents who need aggregated, engagement-weighted briefs from internet sources.
Key capabilities
Evaluates results using engagement signals including upvotes, likes, and real-money odds before an AI judge generates a brief.
Consolidates stories originating from different platforms into unified clusters instead of treating them as separate items.
Limitations and risks
Operation requires internet access and depends on external services including ScrapeCreators and the Perplexity API.
The skill cannot operate without internet access.
Getting started
Install globally using npx skills add mvanhorn/last30days-skill -g, then invoke the skill in a compatible AI agent using /last30days <topic>.