Husky
Run comprehensive CI checks and fix issues until repository is in working state
$ npx claude-code-templates@latest --command="automation/husky" --yesRequires Claude Code. The command adds this command to your project's .claudedirectory — nothing runs on ToolZip's servers.
What's inside this command
Component source
Husky CI Pre-commit Checks
Run comprehensive CI checks and fix issues: $ARGUMENTS
Current Repository State
- Git status: !
git status --porcelain - Package manager: !
which pnpm npm yarn | head -1 - Current branch: !
git branch --show-current - Package.json: @package.json
- Environment file: @.env (if exists)
Task
Verify repository is in working state and fix issues. All commands run from repo root.
CI Check Protocol
Step 0: Environment Setup
- Update dependencies:
pnpm i(unless --skip-install) - Source environment:
.envfile if exists
Step 1: Linting
- Check linter passes:
pnpm lint - Fix formatting issues automatically when possible
Step 2: TypeScript & Build
- Run comprehensive build checks:
pnpm nx run-many --targets=build:types,build:dist,build:app,generate:docs,dev:run,typecheck
- If specific command fails, debug that command individually
- Fix TypeScript errors and build issues
Step 3: Test Coverage
- Source
.envfile first if exists - Run test coverage:
pnpm nx run-many --target=test:coverage - NEVER run normal test command (times out)
- Run individual packages one by one for easier debugging
- For snapshot test failures: explain thesis before updating snapshots
Step 4: Package Validation
- Sort package.json:
pnpm run sort-package-json - Lint packages:
pnpm nx run-many --targets=lint:package,lint:deps
Step 5: Double Check
- If fixes made in any step, re-run all preceding checks
- Ensure no regression introduced
Step 6: Staging
- Check status:
git status - Add files:
git add - EXCLUDE: Git submodules in
lib/*folders - DO NOT COMMIT: Only stage files
Error Handling Protocol
1. Diagnosis
- Explain why command broke with complete analysis
- Cite source code and logs supporting thesis
- Add console logs if needed for confirmation
- Ask for help if insufficient context
2. Fix Implementation
- Propose specific fix with full explanation
- Explain why fix will work
- If fix fails, return to Step 1
3. Impact Analysis
- Consider if same bug exists elsewhere
- Search codebase for similar patterns
- Fix related issues proactively
4. Cleanup
- Remove all added console.logs after fixing
- Run
pnpm run lintto format files - Ask user before staging changes
- Suggest commit message (don't commit)
Development Notes
File Organization
- Functions/types like
createTevmNodeare in:
createTevmNode.js
- Types: TevmNode.ts
- Tests: createTevmNode.spec.ts
Tool-Specific Tips
pnpm i
- If fails, abort unless simple syntax error (missing comma)
pnpm lint (Biome)
- Lints entire codebase
- Auto-fixes most formatting issues
TypeScript Builds
- Look for types in node_modules if not obvious
- For tevm packages, check monorepo structure
- Consult documentation if multiple failures
Test Execution
- Use Vite test runner
- Run packages individually for debugging
- Add console logs to test assumptions
- Explain snapshot changes before updating
Success Criteria
Print checklist at end with ✅ for passed steps:
- ✅ Dependencies updated
- ✅ Linting passed
- ✅ TypeScript/Build passed
- ✅ Tests passed
- ✅ Package validation passed
- ✅ Files staged (no commits made)
Safety Guidelines
- Fix errors proactively - TypeScript/tests will catch regressions
- Never commit - Only stage changes
- One step at a time - Don't proceed until current step passes
- Ask permission before staging if fixes were made
Related Claude Code Commands
Workflow Orchestrator
Orchestrate complex automation workflows with task dependencies, scheduling, and cross-platform execution
Ci Pipeline
Manage and automate CI/CD pipeline configuration with GitHub Actions, multi-environment support, and deployment strategies
Act
Execute GitHub Actions locally using act
Szamlazz
Issue, cancel, and fetch Hungarian invoices via the szamlazz.hu Agent API — with NAV taxpayer lookup and automatic partner cache
Generate Tests
Generate a complete test file for a specified source file or component. Use when the user explicitly asks to write, create, or generate tests for a specific file.
Ultra Think
Multi-framework structured analysis: surfaces hidden assumptions, generates competing solutions, stress-tests each with adversarial reasoning, and delivers confidence-calibrated recommendations
Catalog data and component content are sourced from the open-source davila7/claude-code-templates project (MIT license). ToolZip curates the listing and writes original descriptions; every component links back to its original source. Claude Code is a product of Anthropic. ToolZip is an independent catalog and is not affiliated with or endorsed by Anthropic.