Test

Your Task

Input: $ARGUMENTS

Run automated tests to validate plugin integrity. Execute each test methodically and report results clearly.

Default: Run all tests if no argument provided.

Plugin Test Suite

You are the plugin's automated test runner. Execute each test, track pass/fail, and report actionable results.

Quick Automated Tests (/test quick)

For fast automated validation, run the pytest suite. Call get_python_command() first to get $PYTHON (the venv interpreter path — correct on every OS, including native Windows):

$PYTHON -m pytest ${CLAUDE_PLUGIN_ROOT}/tests/ -v

This covers:

• plugin tests (tests/plugin/) - Frontmatter, templates, references, links, terminology, consistency, config, state, genres, integration
• unit tests (tests/unit/) - State parsers/indexer, shared utilities, mastering functions

Run specific categories:

$PYTHON -m pytest ${CLAUDE_PLUGIN_ROOT}/tests/plugin/test_skills.py -v       # Skills only
$PYTHON -m pytest ${CLAUDE_PLUGIN_ROOT}/tests/plugin/ -v                      # All plugin tests
$PYTHON -m pytest ${CLAUDE_PLUGIN_ROOT}/tests/unit/ -v                        # All unit tests
$PYTHON -m pytest ${CLAUDE_PLUGIN_ROOT}/tests/ -m "not slow" -v               # Skip slow tests

Pytest catches common issues fast. For deep behavioral tests, use the full test suite below.

Output Format

════════════════════════════════════════
CATEGORY: Test Category Name
════════════════════════════════════════

[PASS] Test name
[FAIL] Test name
       → Problem: what's wrong
       → File: path/to/file:line
       → Fix: specific fix instruction

────────────────────────────────────────
Category: X passed, Y failed
────────────────────────────────────────

At the end:

════════════════════════════════════════
FINAL RESULTS
════════════════════════════════════════
config:       X passed, Y failed
skills:       X passed, Y failed
templates:    X passed, Y failed
...
────────────────────────────────────────
TOTAL:        X passed, Y failed, Z skipped
════════════════════════════════════════

TEST CATEGORIES

All test definitions are in test-definitions.md.

14 categories: config, skills, templates, workflow, suno, research, mastering, sheet-music, release, consistency, terminology, behavior, quality, e2e.

Read that file before running tests to understand what each test checks.

RUNNING TESTS

Commands

Quick Tests via Pytest

For rapid validation during development, use pytest directly. Call get_python_command() first to get $PYTHON (the venv interpreter path — correct on every OS):

# Run all tests
$PYTHON -m pytest ${CLAUDE_PLUGIN_ROOT}/tests/ -v

# Run specific test modules
$PYTHON -m pytest ${CLAUDE_PLUGIN_ROOT}/tests/plugin/test_skills.py ${CLAUDE_PLUGIN_ROOT}/tests/plugin/test_templates.py -v

# Verbose with short tracebacks
$PYTHON -m pytest ${CLAUDE_PLUGIN_ROOT}/tests/ -v --tb=short

# Quiet mode (for CI/logs)
$PYTHON -m pytest ${CLAUDE_PLUGIN_ROOT}/tests/ -q --tb=line

Test modules in tests/plugin/:

• test_skills.py - Frontmatter, required fields, model validation
• test_templates.py - Template existence and structure
• test_references.py - Reference doc existence
• test_links.py - Internal markdown links
• test_terminology.py - Deprecated terms check
• test_consistency.py - Version sync, skill counts
• test_config.py - Config file validation
• test_state.py - State cache tool validation
• test_genres.py - Genre directory cross-reference
• test_integration.py - Cross-skill prerequisite chains

Adding New Tests

When bugs are found:

1. Identify which category the test belongs to
2. Add a test that would have caught the bug
3. Run /test [category] to verify test fails
4. Fix the bug
5. Run /test [category] to verify test passes
6. Commit both the fix and the new test

Rule: Every bug fix should add a regression test.

EXECUTION TIPS

• Use Grep with output_mode: content and -n for line numbers
• Use Glob to find files by pattern
• Use Read to check file contents
• Use Bash sparingly (YAML/JSON validation)
• Report exact file:line for failures
• Provide specific, actionable fix instructions
• Group related tests for readability
• Skip tests gracefully if prerequisites missing