Answers are grounded in this paper and cite their sources.
Pre-registration of the second edition of the Public Web Agent Readiness Benchmark (PWARB 2026-II). The study measures whether 200 business-to-business (B2B) companies in six Spanish-speaking Latin American countries (Argentina, Mexico, Colombia, Chile, Peru and Uruguay) expose the public web infrastructure that third-party software agents need in order to discover, interpret and preliminarily evaluate them, and compares them with the ten AI vendors measured in the first edition (March 2026). This document fixes, before data collection: the research questions; two confirmatory hypotheses (H1: fewer than 5% of measurable companies publish an Agent Card at /.well-known/agent-card.json and the median automatic score is below 25 out of 90; H4: technology/SaaS and fintech/banking companies score higher than industry and professional services); the stratified sampling frame and replacement rule; the measurement protocol PARS v2 (100 points, 90 computed automatically from raw HTTP observations, including Agent Card conformance to A2A 1.0, JWS signature verification and endpoint liveness, and 10 double-coded); the collection procedure and ethics rules; the analysis plan; and the handling of deviations. A pilot run of the instrument conducted before registration is disclosed in Section 3. Continues: Marco, G. (2026). Public Web Agent Readiness Benchmark 2026. https://doi.org/10.5281/zenodo.18924996
No takes yet. Share an insight, caveat, or question.
Gabriela Adina Marco (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: