name: log-sentinel
description: rapid diagnostic tool that scans container logs for errors and warnings.

Log Sentinel

Description

This skill quickly retrieves and filters logs from Docker containers or system services to identify errors, warnings, and exceptions. It filters out noise (INFO/DEBUG) to focus on problems.

Usage

Use this skill when the user reports a runtime error, a "500 error", or asks to "check the logs", or uses /scan-logs.

Workflow

1. Identify Target: Determine which service/container to check (e.g., grafana, app_server). If unknown, list running containers.
2. Fetch Logs: Run docker logs --tail 200 <container_id> (or similar).
3. Filter:
   • Keep lines containing: ERROR, WARN, Exception, Fatal, Fail.
   • Discard generic INFO, DEBUG lines unless they provide critical context implies by the error.
4. Report:
   • Present the filtered log lines.
   • If a stack trace or specific error file/line is found, point to the source code file using glance_file or view_file if possible.

Commands

/scan-logs [service_name]

Scans the logs of the specified service.

Reporting

When interpreting the logs or reporting findings to the user, you MUST use Traditional Chinese.