AgentTrust — Pre-Invocation AI Agent Trust & Reliability Checks
Check an AI agent endpoint before you call it: read-only, no API key, returns a trustDecision.
likely compatible
Answered an MCP initialize with no credentials; use auth 'None'.
Rule R2a · what this means · Not tested in Copilot Studio.
| Registry name | io.github.agenttrust-ai/agenttrust |
| Endpoint used for verdict | https://getagenttrust.com/api/mcp/public |
| Transports | streamable-http |
| Auth (Copilot Studio setting) | None (open) |
| Last endpoint check | 2026-10-20T18:14:14Z · initialize OK |
| GitHub | agenttrust-ai/agenttrust |
| Website | https://getagenttrust.com/docs |
| Version | 1.0.4 |
| Registry updated | 2026-10-02 |
Track changes like this weekly. New Copilot-ready MCP servers, breakages and auth changes, in one short email.
Is this wrong or out of date? Email us. Data built 2026-10-20.