From faultline
Shows time-bucketed error occurrence counts from Faultline. Useful for spotting error volume spikes over a configurable window (1h to all).
How this skill is triggered — by the user, by Claude, or both
Slash command
/faultline:statsThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Show error volume over a time window using the `error_stats` MCP tool.
Show error volume over a time window using the error_stats MCP tool.
$ARGUMENTS as the period. Valid values: 1h, 2h, 4h, 1d, 2d, 1w, 1m, all. Default to 1d if empty or invalid.error_stats with the chosen period.bucket | count. Lead with the total and granularity fields from the response so the user knows the resolution./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>.
npx claudepluginhub dlt/faultline --plugin faultlineQueries Faultline for recently active error groups ranked by occurrence count, with a configurable lookback window. Flags brand-new errors and offers drill-down.
Queries Lensflare datasets for recent errors and fatals, groups by frequency, fetches traces, and summarizes root causes with failing spans and services. Use for 'what's broken' or vague stack traces.
Monitors PostHog $exception events for bursts, stuck loops, fingerprint clusters, and status regressions. Emits findings when confidence is high, otherwise records memory and closes out.