description: Show time-bucketed error occurrence counts for the Faultline instance. argument-hint: "[period]" metadata: version: "1.0.1"
Show error volume over a time window using the error_stats MCP tool.
- Parse
$ARGUMENTS as the period. Valid values: 1h, 2h, 4h, 1d, 2d, 1w, 1m, all. Default to 1d if empty or invalid. - Call
error_stats with the chosen period. - Render the response as a short table or sparkline-style summary:
bucket | count. Lead with the total and granularity fields from the response so the user knows the resolution. - If any bucket is dramatically above the others, call it out. Suggest running
/spike for per-group attribution.
This tool is global (across all error groups). For per-group time series, the user should open the group in the dashboard or run /debug <id>.