Claude Idea Debug Skill — Testing skill for Claude Code
Claude Code skill that triggers an IntelliJ IDEA test/debug run from your Claude conversation, captures the log and JUnit results, and pipes them back to Claude for analysis.
How to install Claude Idea Debug Skill
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open Dimoniada/claude-idea-debug-skill and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Claude Idea Debug Skill does
Claude Code skill that triggers an IntelliJ IDEA test/debug run from your Claude conversation, captures the log and JUnit results, and pipes them back to Claude for analysis. Windows only.
Alternatives in Testing
- Debug — You are tasked with helping debug issues during manual testing or implementation 10k ★
- Skill Debug — Debug issues methodically — use when stuck on errors, test failures, or unexpected behavior 2.8k ★
- Fixtests — Systematic test failure analysis and remediation 849 ★
README
idea-debug
A [Claude Code](https://docs.claude.com/en/docs/claude-code/overview) skill that triggers an IntelliJ IDEA test/debug run from your Claude conversation, captures the console log and JUnit results, and pipes them back to Claude for analysis.
**Platform:** Windows only (uses Win32 API + PowerShell + `wscript.exe`).
What it does
You're in Claude Code, you've been editing Java code in IntelliJ. You type:
/idea-debug
Behind the scenes, the skill:
- Brings IntelliJ to the foreground.
- Sends
Debugshortcut (default is Shift+F9, re-run last config — works for both run and debug). - Waits for the test-runner JVM to exit (no timeout, with a 10s heartbeat for long suites).
- Reads the console log file (
debug-capture.log) IntelliJ wrote during the run. - Parses the JUnit XML and prints
[PASS]/[FAIL]per test with stderr. - Minimizes IntelliJ so Claude Code surfaces again — without disturbing your Alt+Tab order.
Claude then summarizes pass/fail and suggests fixes based on the captured stack traces.
Why this exists
Triggering IntelliJ from a Claude Code subprocess on Windows runs into several non-obvious quirks (cross-process focus stealing, JVM daemon reuse, key injection from a non-interactive shell, taskbar focus blocks). This skill works around all of them. See [DESIGN-NOTES.md](#design-notes-the-four-windows-quirks-this-skill-works-around) below for the technical details.
Installation
- Find your Claude Code skills directory (varies by install — search for an existing
skills/folder near your Claude Code data directory). - Copy this repo's contents into a folder named
idea-debuginside that directory:/idea-debug/ ├── SKILL.md └── scripts/ ├── Debug-And-Capture.ps1 ├── Chase-DebugProcess.ps1 └── WinApi.ps1 - Restart Claude Code so it picks up the new skill.
- In IntelliJ, open your project's Run/Debug Configuration and add (one-time setup):
- **Logs t
Related Skills
Six
Use this skill to run test coverage analysis and write tests for uncovered code. Run it when asked to improve
J Debug
Systematic bug investigation — root cause analysis, not random fixes. Use when a test fails, a bug appears, or
Test Notebook
Test the notebook currently being worked on using the notebook-runner subagent and report all results back.
CR Agents
Brute-force AI method for technical work. Personas anchored to experts in their fields run as Claude Code sub-
New Bif Test
Scaffold a JUnit 5 test for a BoxLang BIF. Usage: /boxlang-agent-skills:new-bif-test com.example.bifs.MyFuncti
New Component Test
Scaffold a JUnit 5 test for a BoxLang Component (custom tag). Usage: /boxlang-agent-skills:new-component-test
Related Agents
Failure Intake
Use this agent FIRST when the user brings failing Playwright tests to debug — raw terminal output, a reports/r
Log Reader
Use proactively to read logs, test output, stack traces, CI/CD pipeline output, kubectl/helm/argocd output, PH
Ab Test Analysis
Use when the user wants to analyze A/B test results, interpret p-values, determine statistical significance, o