Show HN: An AI agent runs ten web errands a day, every transcript published

Category: ai-ml

Tags: ai-agents, web-readiness, llms-txt, open-data, crawler-policy, benchmark

Score: 6.3/10 (Innovation: 7, Technical: 5, Documentation: 7, Utility: 6)

Agentability is an open experiment that runs a daily autonomous AI agent through real web errands (e.g., checking kickoff times or earthquake magnitudes) using read-only HTTP and publishes every transcript verbatim, alongside a weekly AI-readiness index scoring the 1219 most-visited sites on conventions like llms.txt, crawler policy, and JSON-LD. Its open data and weekly history make it a useful empirical barometer for the emerging agentic web, though the technical implementation is deliberately simple and the field-test scope is narrow.

Target audience: backend devs, web developers, SEO/DevRel teams, AI researchers, devops

Repository: https://agentability.org/ · TypeScript · Apache-2.0 · 6 stars

View on Hacker News