feat(skills): Refactor tmux-testing to claude-isolated-test with progressive disclosure

[baleen37]

Summary

Refactor tmux-testing skill into claude-isolated-test with comprehensive progressive disclosure structure and focus on fresh instance testing methods.

Key Changes

• Renamed skill: tmux-testing → claude-isolated-test for broader scope
• Reduced SKILL.md: From 961 lines (~7,000 words) to ~370 lines (1,321 words)
• Progressive disclosure: Separated content into core patterns, detailed references, and working examples
• Four fresh instance methods: tmux sessions, Docker/packnplay, git worktrees, environment variables
• Improved description: Third-person format with specific trigger phrases (284 chars vs 374 chars)
• Standardized writing: Consistent imperative/infinitive form throughout

Structure

claude-code-testing/
├── SKILL.md                          # Core patterns, quick reference (1,321 words)
├── examples/                         # Production-ready bash scripts
│   ├── tmux-fresh-instance.sh       # Basic tmux testing
│   ├── sessionstart-test.sh         # SessionStart/Stop hook testing
│   └── parallel-testing.sh          # Parallel configuration testing
├── references/                       # Detailed documentation
│   └── advanced-testing.md          # Hook patterns, debugging workflows
└── scripts/                          # Reusable utilities
    └── wait-for-output.sh           # Polling helper function

Quality Improvements

• ✅ All bash scripts use proper traps and error handling
• ✅ Executable permissions set correctly
• ✅ Reduced duplication with shared utilities
• ✅ Skill-reviewer validated: Pass with minor polish applied
• ✅ Follows plugin-dev best practices exactly

Testing

Validated with skill-reviewer agent:

• Description quality: ✅ Specific trigger phrases, third-person format
• Content organization: ✅ Progressive disclosure implemented
• Writing style: ✅ Imperative form standardized
• Working examples: ✅ All scripts production-ready

Test plan

• Skill loads correctly with updated name
• Trigger phrases work as expected
• Example scripts execute without errors
• Progressive disclosure flows naturally

🤖 Generated with Claude Code

Summary by CodeRabbit

• New Features

  • Added testing examples and reusable automation scripts for isolated test environments.
  • Enabled parallel testing capabilities with automated startup handling and real-time output monitoring.

• Documentation

  • Added comprehensive testing patterns, hook configurations, debugging workflows, and best practices.
  • Consolidated testing skill documentation into unified guidance for developers.

[coderabbitai]

Caution

Review failed

The pull request is closed.

📝 Walkthrough

Walkthrough

The changes introduce a new comprehensive "Claude Code Isolated Testing" skill containing documentation, example scripts, and reusable utilities for testing Claude Code using tmux, Docker, Git worktrees, and environment-variable-based isolation methods. The previous tmux-testing skill documentation is replaced with expanded content.

Changes

Cohort / File(s)	Summary
Claude Code Testing Skill Documentation plugins/me/skills/claude-code-testing/SKILL.md, plugins/me/skills/claude-code-testing/references/advanced-testing.md	Added comprehensive documentation defining the Claude Code Isolated Testing skill with multiple isolation methods, testing patterns, debugging workflows, hook types, MCP integration guidance, and known issues. Includes feature coverage for Claude Code 2.1.0+ (hot reload, frontmatter hooks, scoped hooks, forked sub-agent context).
Testing Example Scripts plugins/me/skills/claude-code-testing/examples/parallel-testing.sh, plugins/me/skills/claude-code-testing/examples/sessionstart-test.sh, plugins/me/skills/claude-code-testing/examples/tmux-fresh-instance.sh	Added three Bash scripts orchestrating isolated Claude Code testing via tmux: parallel testing across multiple instances with trust-prompt handling, SessionStart/SessionStop hook verification, and fresh single-instance testing workflows.
Testing Utilities plugins/me/skills/claude-code-testing/scripts/wait-for-output.sh	Added reusable polling helper function for tmux session output verification with configurable timeout and error reporting.
Removed Documentation plugins/me/skills/tmux-testing/SKILL.md	Deleted obsolete tmux-testing skill documentation, subsumed by the broader claude-code-testing skill.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~15 minutes

Possibly related PRs

• feat(me): add tmux-testing skill for TTY-required hook testing #436: Directly replaces and expands the previous tmux-testing skill documentation with the new claude-code-testing skill that subsumes tmux-based testing guidance.
• feat(strategic-compact): plugin improvements with session-based state and tmux-testing skill #214: Adds tmux-based testing documentation and scripts that exercise SessionStart hooks and session-environment behaviors, complementing the testing workflows now documented here.

Poem

🐰 Hops through tmux with glee,
Testing Claude in harmony,
Sessions bloom, hooks align,
Parallel tests run so fine,
Better docs to guide the way,
Skill complete—hip-hip-hooray! 🎉

✨ Finishing touches

• 📝 Generate docstrings

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch feat/tmux-claude-test

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}