AI research & benchmarks / Agentic web usability benchmark
An open public experiment that tests whether AI agents can complete everyday web errands, publishes the full transcripts, and scores 113 sites for agent-readiness.
The page clearly presents a real AI-related product/project: a weekly public benchmark and observatory for agent web usability. It shows concrete mechanics, published transcripts, site scoring, open data, reproducible checks, and enough detail to understand how the project works and why it matters.
The snapshot explains the experiment well, but it does not show the underlying data model, scoring rubric in detail, or exactly which checks are used beyond broad categories like llms.txt, crawler policy, readable content, structured data, and MCP.
Last evaluated just now. Current rank #32. Down 2 spots in the rankings.
No related listings yet.
