Jin - AI Agent Protocol

An open-source protocol that makes the internet machine-readable for AI agents, eliminating the need for expensive scraping infrastructure. Allows webmasters and agent builders to define intents so AI can easily extract what it needs without breaking websites.

The push toward agentic AI workflows in 2024-2025 has created real demand for structured ways to expose web content to LLM-based agents, since most sites are still built for humans and scraping pipelines break constantly. The closest analogy is robots.txt or structured data schemas like Schema.org, but there's no clear incumbent owning the "agent-readable web" protocol layer yet. The $2k-10k/mo revenue band is plausible only if monetization comes from hosted tooling or managed services on top of the open-source core, since protocols themselves are notoriously hard to charge for directly. The biggest risk is adoption: a protocol is worthless without simultaneous buy-in from both webmasters and agent builders, and that two-sided cold-start problem has killed better-funded standardization efforts before.