A focused Sentinel sweep across your MCP servers: reachability, initialize handshake, tools/list, latency, schema drift, and a prioritized remediation report.
Sentinel's public CLI can fail a release when an MCP server is down or degraded. Version 0.1.9 preserves JSON evidence before exit and turns that saved result into the first draft of the paid reliability audit report.
npx --yes @megamcp/sentinel@latest check https://your-server.example/mcp --fail-on-degraded --latency-budget-ms 5000 --min-tools 3 --output sentinel-result.jsonnpx --yes @megamcp/sentinel@latest audit sentinel-result.json --output mcp-reliability-audit.mdUse --header for bearer-token targets. Attach sentinel-result.json or mcp-reliability-audit.md to the audit request so the paid review starts from the same protocol evidence.
It is designed for operators who need evidence before they buy monitoring or route production traffic through a gateway.
Sentinel verifies MCP reachability, initialize, tools/list, and latency instead of relying on a generic HTTP ping.
The report flags added, removed, and changed tools so teams can see contract risk before agents break.
You get ranked findings, target-level status, and next actions that can roll into Team monitoring.
The audit is a cash-now service wedge for Sentinel. The $499 fee is credited toward three months of Team monitoring when the account converts. When online payment is enabled, checkout uses a Stripe-hosted link.