ashworks1706/Decoy

Decoy is a classic hackathon project (winner of a Google-sponsored event) that demonstrates the 'AI Honeypot' concept. While the vision of using agents to interact with scammers for intelligence gathering is conceptually sound, the project lacks any meaningful market signal with only 2 stars and 0 forks after six months. From a technical perspective, RAG-based security pipelines are now common architectural patterns rather than unique IP. The defensibility is nearly non-existent as the project lacks a proprietary dataset, a significant user base, or a complex technical moat. Frontier labs and established security vendors (e.g., Abnormal Security, Ironscales, or Microsoft Defender) are already deploying much more sophisticated agentic workflows for social engineering detection. This project serves as a good reference for how to combine RAG with agents for security, but it is unlikely to survive as a standalone entity against platform-level AI integrations from Google or Microsoft.