{"id":"io.tooloracle/agentguard","name":"agentguard","description":"AgentGuard — 20-tool AI safety MCP: policy preflight, risk scoring, audit logging, rate limits.","connect":{"transport":"streamable-http","url":"https://feedoracle.io/guard-oracle/mcp/"},"website":"https://github.com/ToolOracle/agentguard","version":"1.0.0","source":"official MCP registry (registry.modelcontextprotocol.io)","reliability":{"verdict":"failing","attempts":3,"successes":0,"successRate":0,"consecutiveSuccesses":0,"lastStatus":"gone","lastCheckedAt":"2026-11-03T06:15:37.822Z","lastLatencyMs":97,"toolsAdvertised":null,"method":"MCP initialize + tools/list from Faceabot, identified as FaceabotProbe/1.0 plus, when the server declares a read-only tool (readOnlyHint), one call of that tool with a neutral sample input to check it really returns a result. Servers on the 2026-07-28 MCP revision (no initialize handshake) get the same checks in that revision","work":{"verdict":"not_tested","note":"Faceabot has not run a real task on this server yet (it only tests tools that declare themselves read-only)."}},"details":"https://faceabot.com/api/catalog?id=io.tooloracle%2Fagentguard","page":"https://faceabot.com/tools/io.tooloracle/agentguard","badge":{"image":"https://faceabot.com/badge/io.tooloracle/agentguard.svg","markdown":"[![Measured by Faceabot](https://faceabot.com/badge/io.tooloracle/agentguard.svg)](https://faceabot.com/tools/io.tooloracle/agentguard)","note":"Live reliability badge the publisher of this server can show on its README or website."},"recentChecks":[{"at":"2026-11-03T06:15:37.822Z","ok":0,"status":"gone","ms":97,"f_status":null},{"at":"2026-10-30T15:35:39.708Z","ok":0,"status":"gone","ms":724,"f_status":null}]}