MCP server · operational

agentcheck

Monitor a production AI agent or MCP server for drift and regressions: enroll a target, add scheduled checks judged against goldens, run them now, record traces and promote one to a replayable check, read uptime, incidents and pricing.

agentcheck watches production AI agents and MCP servers and re-runs your checks on a schedule, judging each answer against a golden with an LLM judge where needed. It is quiet by default: it contacts you only when something changed. Public per-target surfaces need no key: a status page, a badge SVG and a trust.json with uptime, last nightly scores and adversarial pass rate. Read tools (agentcheck_get_status, agentcheck_get_pricing) need no key. Write tools need a Bearer API key that an agent can issue itself with POST https://agentwares-agentcheck.vercel.app/api/v1/signup {"accept_terms": true}; no email, no CAPTCHA. Errors are JSON with code, cause, fix and retryable. The free tier monitors one target hourly; paid tiers add targets, 5-minute checks, nightly suites and trace-to-check. Also publishes a monthly liveness census of the MCP registry at https://agentwares-mcp-liveness.vercel.app and a public model drift index at https://agentwares-agentcheck.vercel.app/drift.json.

endpoint: https://agentwares-agentcheck.vercel.app/api/mcp
auth: none for read tools (status, pricing); Bearer API key for write tools, issued without a human by POST /api/v1/signup
try it: tested example call
curl -s -X POST https://agentwares-agentcheck.vercel.app/api/mcp -H 'content-type: application/json' -H 'accept: application/json, text/event-stream' -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"agentcheck_get_pricing","arguments":{}}}'
capabilities: monitor-mcp-server · monitor-ai-agent · check-uptime · detect-drift · detect-regression · run-checks-now · record-trace · promote-trace-to-check · list-incidents · get-status · get-pricing · create-monitor-target
tags: monitoring · observability · mcp · ai-agents · testing
pricing: Free tier (1 target, hourly checks, badge, status page) then $29/month for 3 targets at 5-minute checks
i/o: application/json (JSON-RPC 2.0 over Streamable HTTP) → application/json
docs: https://agentwares-agentcheck.vercel.app/llms.txt
machine-readable: /api/public/registry/agentcheck-mcp

Get a key in one call

No account, no email, no dashboard

100 calls/day anonymous · 1,000/day with this free key · 50,000/day on Agent Pro. The key is returned instantly in the response — send it as the x-api-key header on any /api/public/* request.

tested example call

curl -s -X POST https://agentnexus.app/api/public/keys \
  -H 'content-type: application/json' \
  -d '{"agent":"my-agent","purpose":"tool discovery"}'

Use it from your agent

Point any MCP-capable agent (Claude, Cursor, your own) at the no-auth endpoint — it can then find agentcheck and the rest of the registry on its own:

tested example call

claude mcp add --transport http agent-nexus https://agentnexus.app/api/public/mcp

Agents can discover agentcheck through the Agent Nexus MCP server, /llms.txt or the discovery API.