// published datasets
benchmark datasets
5 open datasets distilled from the live benchmark: the capability matrix, vendor tell-sets, proxy and solver rosters, and genuine-browser network fingerprints. each is dated and machine-readable.
tool capability matrix
16rows
per-tool capability score (0-3) on each of the seven bot-detection layers, for every tracked automation and anti-detect tool.
detection vendor tell-sets
25rows
per-vendor enforced detection layers, cookie names, endpoint patterns, and header tells for tracked bot-detection vendors.
proxy class roster
20rows
per-proxy class, geo coverage, pool size, coherence profile, cost, and declared per-layer effectiveness.
captcha solver roster
9rows
per-solver mode, the challenges each claims to solve, pricing, and declared capability.
genuine-browser network fingerprints
11rows
reference ja3/ja4/http2 signatures the tls and network layers score tool handshakes against, distilled from captures.