tests.yml

name: Tests

on:
  push:
    branches: [ main ]
  pull_request:
    branches: [ main ]

jobs:
  test:
    runs-on: ubuntu-latest
    strategy:
      matrix:
        python-version: ["3.10", "3.11", "3.12"]

    steps:
    - uses: actions/checkout@v4

    - name: Set up Python ${{ matrix.python-version }}
      uses: actions/setup-python@v5
      with:
        python-version: ${{ matrix.python-version }}

    - name: Install core package (no external deps)
      run: pip install -e ".[dev]"

    - name: Run zero-dep proof
      run: python verify_moa.py

    - name: Run pytest suite
      run: pytest tests/ -v --tb=short

    - name: Run safety benchmarks (jailbreak + FP check)
      run: |
        python -c "
        import sys, pathlib
        sys.path.insert(0, str(pathlib.Path('.').resolve()))
        from benchmarks.bench_safety import run_jailbreak_battery, run_false_positive_test
        total, blocked = run_jailbreak_battery()
        assert blocked == total, f'JAILBREAK FAILURE: {total-blocked} attacks succeeded'
        fp_total, fp_blocked = run_false_positive_test()
        assert fp_blocked == 0, f'FALSE POSITIVE: {fp_blocked}/{fp_total} legitimate actions blocked'
        from benchmarks.bench_jailbreak import run_jailbreak_suite
        r = run_jailbreak_suite()
        assert r.block_rate == 1.0, f'ADVERSARIAL: {r.bypassed} attacks bypassed'
        print(f'Safety benchmarks: {blocked}/{total} jailbreaks blocked, 0 false positives, {r.total} adversarial blocked')
        "