Analyze structured NDJSON wide-event logs from evlog file system drains.
npx skills add https://github.com/jasonviipers/vipers --skill analyze-logs-jasonviipers
Please help me install this Agent Skill. Skill: analyze-logs Source: https://github.com/jasonviipers/vipers/tree/main/.agents/skills/analyze-logs Command: npx skills add https://github.com/jasonviipers/vipers --skill analyze-logs-jasonviipers
error.message, error.data.why, and error.data.fix to explain failures and suggest remediation. - Performance Analysis: Filter and sort on durationMs to find slow endpoints, degrading performance, and time-of-day patterns. - Request Tracing & Patterns: Trace a single request by requestId, compute error rates per endpoint, and compare client-side versus server-side errors. - Use Case: A user reports intermittent 500 errors on the checkout endpoint. The Skill reads the latest .jsonl log file, filters error events for /api/checkout, groups them by error.message, and surfaces the root cause with the suggested fix. ## Quick Start Analyze the logs in .evlog/logs and tell me why the recent requests to /api/checkout are failing.High-intent search queries and answers about installing and using this skill.
Read the most recent .jsonl file in .evlog/logs/, parse each line as a JSON event, and filter for level "error" or status >= 400. Each event contains error.message, error.data.why, and error.data.fix fields that explain the failure and suggest remediation.
Filter events where durationMs exceeds your threshold (e.g. 1000) and sort descending. Group by path to identify slow endpoints and check timestamps for time-of-day patterns.
Evlog writes one self-contained JSON wide event per line in NDJSON mode, or multi-line indented JSON in pretty mode. Files are named by date, like 2026-03-14.jsonl, and each event includes timestamp, level, method, path, status, durationMs, and requestId.
The file system drain is not enabled. Run npx @evlog/cli init --yes --drain fs to wire it automatically, or manually pass createFsDrain() from evlog/fs to your framework's evlog integration, then generate traffic to produce logs.
Yes. Filter events by the requestId field to retrieve the complete wide event for that request. Unlike traditional logs, one line contains all context for the request, so no multi-line correlation is needed.
zhcndocExtract and interpret NDJSON log events from .evlog/logs/ to diagnose errors and evaluate performance.
ESO-ToolkitDebug failed GitHub Actions CI runs by identifying the failed stage and root cause.
whawkinsivAutomate uptime, error tracking, and performance monitoring for web apps.
vertivolatamConfigure structured logging and centralized log aggregation with Loki, Elasticsearch, Fluentd, and Promtail.