Agent Skills › DetachHead/rebased › debugging

debugging

GitHub

提供 IntelliJ IDEA 调试技巧，包括 idea.log 搜索模式及端到端问题排查方法，用于分析日志、定位 IDE 故障和异常。

.claude/skills/debugging/SKILL.md DetachHead/rebased

Trigger Scenarios

IDEA 启动卡顿或冻结需要分析 idea.log 中的特定类日志排查用户入口行为不符合预期的问题调查代码执行链路中的异常

Install

npx skills add DetachHead/rebased --skill debugging -g -y

More Options

Non-standard path

npx skills add https://github.com/DetachHead/rebased/tree/master/.claude/skills/debugging -g -y

Use without installing

npx skills use DetachHead/rebased@debugging

指定 Agent (Claude Code)

npx skills add DetachHead/rebased --skill debugging -a claude-code -g -y

安装 repo 全部 skill

npx skills add DetachHead/rebased --all -g -y

预览 repo 内 skill

npx skills add DetachHead/rebased --list

SKILL.md

Frontmatter

{
    "name": "debugging",
    "description": "Debugging techniques and tools for IntelliJ development. Use when debugging IDE issues or investigating failures."
}

Debugging

Documentation

Logger FAQ - Frequently asked questions about logging
How to Investigate Freezes - Diagnosing IDE freezes
Exception Analyzing FAQ - Analyzing and categorizing exceptions

Searching idea.log

When debugging IntelliJ code, search idea.log using these patterns:

Log Levels:

FINE = debug level (from LOG.debug { })
INFO = info level (from LOG.info())

Category Format: #<abbreviated.package.path>.<ClassName>

Example: com.intellij.openapi.wm.impl.status.IdeStatusBarImpl → #c.i.o.w.i.s.IdeStatusBarImpl

Search Examples:

# Debug logs from IdeStatusBarImpl
grep "FINE - #c.i.o.w.i.s.IdeStatusBarImpl" idea.log

# Info logs from IdeStatusBarImpl
grep "INFO - #c.i.o.w.i.s.IdeStatusBarImpl" idea.log

# All logs from a class
grep "#c.i.o.w.i.s.IdeStatusBarImpl" idea.log

End-to-End Behavior

When fixing issues where a user-facing entry point (script, command, API) doesn't behave as expected:

Trace the full execution chain from entry point to actual executor before making changes
Identify all layers that handle the relevant behavior - each layer may swallow, transform, or ignore it
Verify the fix at the entry point the user actually invokes, not just at the layer you modified
Add debugging output if still unsure - add to all layers to trace the issue