home / skills / aidotnet / moyucode / log-analyzer

log-analyzer skill

safe

This skill analyzes log files, enabling pattern search, filtering, statistics, and error detection to improve debugging and operational insights.

npx playbooks add skill aidotnet/moyucode --skill log-analyzer

Review the files below or copy the command above to add this skill to your agents.

Files (2)

SKILL.md

965 B

---
name: log-analyzer
description: 解析和分析日志文件，支持模式匹配、过滤、统计和错误检测。
metadata:
  short-description: 分析日志文件
source:
  repository: https://github.com/logpai/logparser
  license: MIT
---

# Log Analyzer Tool

## Description
Parse and analyze log files to extract patterns, filter entries, generate statistics, and detect errors.

## Trigger
- `/logs` command
- User needs to analyze logs
- User wants to find errors in logs

## Usage

```bash
# Analyze log file
python scripts/log_analyzer.py app.log

# Filter by level
python scripts/log_analyzer.py app.log --level ERROR

# Search pattern
python scripts/log_analyzer.py app.log --grep "connection failed"

# Get statistics
python scripts/log_analyzer.py app.log --stats

# Tail mode
python scripts/log_analyzer.py app.log --tail 100
```

## Tags
`logs`, `analysis`, `debugging`, `monitoring`, `errors`

## Compatibility
- Codex: ✅
- Claude Code: ✅

Overview

This skill parses and analyzes log files to extract patterns, filter entries, produce statistics, and detect errors. It provides command-style controls for level filtering, pattern searches, tailing, and summary statistics to speed debugging and monitoring. Designed for TypeScript-based environments but useful for any plain-text logs.

How this skill works

The tool reads log files and applies user-specified operations: pattern matching (grep-style), level filtering (e.g., ERROR), and real-time tailing. It computes simple statistics (counts by level, frequency of messages, time-window summaries) and flags likely errors or anomalous spikes. Results are printed to the console or streamed for pipelines.

When to use it

Investigating application crashes or repeated errors
Filtering large logs to focus on specific severity levels
Searching for occurrences of a text or regex pattern
Generating quick statistics about traffic or error rates
Tailing logs in real time during deployments or troubleshooting

Best practices

Run pattern searches with precise regex to reduce noise
Combine level filters with time ranges when available to target relevant windows
Use the --stats summary before deep inspection to prioritize investigation
Pipe filtered output to other tools (grep, awk, jq) for complex analysis
Keep a copy of raw logs before aggressive filtering to preserve auditability

Example use cases

Run a full analysis to enumerate ERROR occurrences and top error messages from a crash window
Tail the last 100 lines of a service log during a deploy to watch for regressions
Filter by level ERROR to create a short report for an on-call rotation
Search logs for 'connection failed' across multiple files to identify network issues
Generate statistics to compare error rates before and after a code change

FAQ

What input formats are supported?

Plain-text log files with line-based entries. Structured JSON logs are supported if lines contain parseable JSON objects.

Can I use regex patterns?

Yes. The search option accepts regular expressions for flexible pattern matching.