Live health, breaking-change and spec-readiness checks for remote MCP servers in the official registry. Read-only probes every 15 minutes.
| Server | Uptime | Latency | Tools | Checks |
|---|---|---|---|---|
com.ufcalendar/fight-api api.ufcalendar.com/mcp | 100% | 169 ms | 47 | 6 |
fun.oaia/arena api.oaia.fun/mcp | 100% | 375 ms | 7 | 6 |
com.moltlinestudio/dropship mcp.moltlinestudio.com/dropship | 100% | 98 ms | 6 | 5 |
com.moltlinestudio/shopify mcp.moltlinestudio.com/shopify | 100% | 112 ms | 6 | 5 |
cloud.dchub/mcp-server dchub.cloud/mcp/grok | 100% | 185 ms | 11 | 5 |
com.kythene/kythene kythene.com/mcp/kythene | 100% | 415 ms | 48 | 5 |
com.flat-hunt/mcp mcp.flat-hunt.com/mcp | 100% | 456 ms | 11 | 5 |
dev.rumoro/mcp mcp.rumoro.dev/mcp | 100% | 779 ms | 55 | 5 |
com.techsnif/mcp mcp.techsnif.com | 100% | 46 ms | 13 | 4 |
com.moltlinestudio/shipping mcp.moltlinestudio.com/shipping | 100% | 62 ms | 6 | 4 |
com.moltlinestudio/data mcp.moltlinestudio.com/data | 100% | 67 ms | 7 | 4 |
cloud.dchub/mcp-server dchub.cloud/mcp/claude | 100% | 156 ms | 67 | 4 |
| Server | Error |
|---|---|
com.a2awire/benchmark-predict-the-news a2awire.com/mcp/benchmarks/predict-the-news/http | timeout |
io.github.tjcgraham-rgb/gaip-trust-assurance www.gaipagents.com/mcp/agents/trust-assurance | timeout |
com.theresanaiforthat/taaft theresanaiforthat.com/mcp-server/ | timeout |
io.github.ogasurfproject-jpg/hs-partner-002-mcp p002.horizonshield.dev/mcp | ConnectTimeout |
io.github.reefapi/reefapi-mcp api.reefapi.com/mcp | timeout |
com.aigramx/aigramx aigramx.com/api/v1/mcp | timeout |
io.github.AgentMindCloud/agent-search-pro aggregator-beta.vercel.app/mcp | HTTP 404 |
to.agentservices/agentservices agentservices.to/mcp | HTTP 503 |
io.github.gmannerHJ/agent-hub agents.nomoe.ai/mcp | ConnectError |
io.github.yyxqqq777/agentpay agentpay-xhs-production.up.railway.app/mcp | HTTP 404 |
How often a verdict still held at the very next scheduled check of the same server (usually ~15 minutes later). This measures how stable our verdicts are between checks, not whether real tool calls succeed. History starts Oct 15, 2026, so samples are still small. Raw data
1,000 checks/day and 3 change-alert watches. Instant, no email or signup. Agents can also call POST /v1/keys or the get_api_key MCP tool.
curl -H "X-API-Key: $TW_KEY" \
"https://toolweather.toolweather.workers.dev/v1/is_safe_to_call?server=com.ufcalendar/fight-api"
{ "verdict": "safe", "status": "ok", "latency_ms": 182,
"spec_readiness": {"readiness": "dual", "supports_2026_07_28": true},
"last_tool_change": null, "last_checked": "…" }
# or add the MCP server and call the tool:
is_safe_to_call(server="com.ufcalendar/fight-api", tool="search")agentbit.browser_renderdemand.requestaudio.transcribeaudio.transcribemarket.quotemarket.quotecatalog.searchstanding_ways[](https://toolweather.pages.dev/)