Card snapshot

This is a frozen copy of the agent's agent-card.json as we observed it at the timestamp above. We capture a new snapshot every time the card's content hash changes. Useful for: forensic drift analysis, verifying downstream callers see the right version, reproducing routing decisions made historically.

{
  "name": "Known Good",
  "description": "Agentic search engine for AI agents: an index of websites verified usable by agents \u2014 every capability probed and tested live, tier 1\u20133 (declared \u00b7 probe-verified \u00b7 behaviourally verified), dated and citable. find_capability answers \"who can do X\"; get_site_report answers \"is this specific host agent-ready\", returning the dated evidence for one site. Nothing is hand-picked and placement cannot be bought: an automated probe decides. Human index of working MCP servers: https://knowngood.sh/mcp-servers",
  "version": "1.7.0",
  "provider": {
    "organization": "Known Good",
    "url": "https://knowngood.sh"
  },
  "documentationUrl": "https://knowngood.sh/api",
  "iconUrl": "https://knowngood.sh/favicon.svg",
  "supportedInterfaces": [
    {
      "url": "https://knowngood.sh/a2a",
      "protocolBinding": "JSONRPC",
      "protocolVersion": "1.0"
    }
  ],
  "capabilities": {
    "streaming": false,
    "pushNotifications": false,
    "extendedAgentCard": false
  },
  "defaultInputModes": [
    "text/plain"
  ],
  "defaultOutputModes": [
    "text/plain",
    "application/json"
  ],
  "skills": [
    {
      "id": "check-webmcp",
      "name": "Check a site for WebMCP tools",
      "description": "Check a website LIVE for WebMCP tools, and where a browser is available whether any tool actually registers when the page runs. Returns one of: registered, declared_only, declared, not_declared, could_not_ask, or excluded when the site's robots.txt refuses AI input. could_not_ask is a distinct answer and is never reported as a no. Tools are listed, never called. Reads only the site named plus published population figures \u2014 it does not consult the index, and running it changes no ranking or listing.",
      "tags": [
        "webmcp",
        "verification",
        "live-check"
      ],
      "examples": [
        "example.com",
        "https://docs.example.com/"
      ],
      "inputModes": [
        "text/plain"
      ],
      "outputModes": [
        "text/plain",
        "application/json"
      ]
    },
    {
      "id": "find-capability",
      "name": "Find agent-ready websites",
      "description": "Find websites where an agent can accomplish a task. Returns probe-verified agent-readiness signals, unverified apparent actions read from page content, the verification tier, how each site was found (it declared agent access, probe discovery, or it was submitted), and a report URL to cite. Returns count:0 with an empty_note when nothing in the index matches \u2014 never a nearest guess.",
      "tags": [
        "search",
        "discovery",
        "agent-readiness",
        "verification"
      ],
      "examples": [
        "buy perfume",
        "book scuba diving",
        "currency conversion API",
        "seo agency"
      ],
      "inputModes": [
        "text/plain"
      ],
      "outputModes": [
        "text/plain",
        "application/json"
      ]
    },
    {
      "id": "get-site-report",
      "name": "Fetch one site's verification report",
      "description": "Fetch Known Good's dated verification report for one listed website: which capabilities were tested, which passed, when it was probed, what an agent can apparently do there, and how the site was found: whether it declared agent access, came from probe discovery, or was submitted. Answers \"is this specific host agent-ready?\". Returns found:false for anything not in the index \u2014 it never guesses a nearest match. Cite the report URL and its date.",
      "tags": [
        "verification",
        "site-report",
        "evidence",
        "provenance"
      ],
      "examples": [
        "developers.cloudflare.com",
        "www.siemens.com"
      ],
      "inputModes": [
        "text/plain"
      ],
      "outputModes": [
        "text/plain",
        "application/json"
      ]
    }
  ]
}