Answers are grounded in this paper and cite their sources.
Evaluates public web infrastructure readiness for autonomous agents, indicating maturity gaps in readiness.
This paper introduces Public Web Agent Readiness as a measurable construct for evaluating whether an organization's public web infrastructure is sufficient for third-party autonomous agents to discover, interpret, and preliminarily evaluate it. The study presents the PARS (Public Agent Readiness Score), a 100-point benchmark protocol comprising a Core Standards Score (70 points) and an Ecosystem Signals Score (30 points), grounded in RFC 8615, RFC 9309, and the Agent2Agent (A2A) discovery model. This benchmark is an empirical companion to Beyond AEO: The AI Visibility Stack and the Era of Agent Discovery Optimization (Marco, 2026; https://doi.org/10.5281/zenodo.18728630), applying its Layer 2 (Agent Discoverability) framework to a purposive sample of ten leading AI-forward vendors: OpenAI, Anthropic, Salesforce, ServiceNow, Google Cloud, Zapier, UiPath, Automation Anywhere, AWS, and IBM. The principal finding is that the sampled market is significantly more mature in agent productization than in public root-level agent discovery infrastructure. No vendor in the sample exposed a confirmed, directly retrievable Agent Card at the current standard well-known path (/.well-known/agent-card.json). Scores ranged from 32 to 59 out of 100, with all cases falling in the "Emerging / Partial" or "Low Observable Readiness" tiers. The deposit includes the full working paper and the primary dataset (PWARB_dataset_v1.csv) with CSS, ESS, PARS scores, tier classification, and key observations for all ten cases.
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: