lanes: Skill for Claude Code

Implement and Run Tests

Use the project's Mocha and VS Code test setup to verify task requirements.

Your Workflow

1. Read the Test Plan

Read tests.json to understand what tests need to be implemented. The coder skill creates this file with planned tests. For a standalone testing request without an existing plan, derive the cases from the task requirements first:

{
  "planned": [
    {
      "id": "test-id",
      "description": "What the test verifies",
      "file": "src/test/session/session-provider.test.ts",
      "suite": "Suite name",
      "priority": "critical|high|medium|low",
      "acceptance_criteria": ["Given X, when Y, then Z"],
      "implemented": false
    }
  ]
}

2. Implement Tests

For each test in tests.json (ordered by priority: critical → high → medium → low):

1. Read the target test file specified in file
2. Implement the test according to the acceptance_criteria
3. Follow existing test patterns in the file
4. Ensure proper setup/teardown
5. Mark the test as "implemented": true in tests.json

3. Run Tests

After implementing:

npm test

Verify all tests pass.

4. Clean Up

When all tests are implemented and passing:

• Delete tests.json

Test Implementation Constraints

1. Mocking: Mock cp.exec and fs.promises to avoid real file system changes in unit tests. Use temp directories for integration tests.

2. Async: Ensure tests await extension activation before checking command registration. The extension activates lazily.

3. Isolation: Verify terminals and worktrees are cleaned up after tests. Use setup() and teardown() hooks.

4. Prioritization: Implement critical priority tests first, then high, medium, low.

Project Test Setup

• Tests are in src/test/, mirroring the source structure
• Test runner: @vscode/test-cli with Mocha
• Config: .vscode-test.mjs
• Tests run in a real VS Code instance via @vscode/test-electron
• Use fs.mkdtempSync() for temporary test directories
• Clean up with fs.rmSync(dir, { recursive: true, force: true })