detect_anomalies
Scan tables for unusual patterns: volume changes, data gaps, high null rates, and stale data. Detect anomalies with severity-ranked alerts. Automatically samples large tables for efficiency.
Instructions
Scan a table for unusual patterns: volume drops/spikes, data gaps, value concentration, high null rates, stale data. Severity-ranked alerts. Tables > 100k rows use a sampled path (~5%). Dialect-aware sampling.
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Implementation Reference
- bin/server.js:52-57 (registration)Registration of detect_anomalies tool in the static TOOLS array for the local stdio adapter. This is a stub/redirect adapter — no actual handler logic exists in this codebase.
{ name: "detect_anomalies", description: "Scan a table for unusual patterns: volume drops/spikes, data gaps, value concentration, high null rates, stale data. Severity-ranked alerts. Tables > 100k rows use a sampled path (~5%). Dialect-aware sampling.", inputSchema: { type: "object" }, }, - bin/server.js:190-193 (handler)Generic catch-all CallTool handler that returns a redirect message for ALL tools including detect_anomalies. The actual execution happens on the remote hosted server at https://data.thinair.co/mcp.
server.setRequestHandler(CallToolRequestSchema, async () => ({ content: [{ type: "text", text: REDIRECT_MESSAGE }], isError: false, }));