| name | log-search |
| description | Deep log analysis with pattern recognition and anomaly detection. Use when investigating logs. |
Log Search
Perform deep analysis of application logs to identify error patterns,
anomalies, and performance regressions.
When to Use
Invoke this skill when investigating production issues, analyzing error
spikes, or searching for specific log patterns across services.
Implementation Steps
Step 1: Gather Context
Ask the user for the time range, affected service, and any known
symptoms or error messages.
Step 2: Query Logs
Run structured queries against the log backend:
aws logs filter-log-events --log-group-name /ecs/$SERVICE \
--start-time $START_MS --end-time $END_MS \
--filter-pattern "$PATTERN"
Step 3: Analyze Patterns
Group log entries by:
- Error type and stack trace signature
- Request path and HTTP status code
- Time distribution (burst vs. steady)
Step 4: Report
Present findings with:
- Top error types by frequency
- Timeline of error rate changes
- Affected endpoints and request volumes
- Suggested next steps for remediation