Auto-Trader/.github/workflows/test.yml

name: CI

on:
  push:
    branches: [main]
  pull_request:
    branches: [main]
  # Manual trigger. On 2026-08-06 twelve commits landed on main without a
  # single push-event run being created, and there was then no way to get a
  # verdict on main short of pushing an empty commit.
  workflow_dispatch:

# `npm ci` runs a security audit after installing and blocks on the response.
# npmjs.com's advisory endpoint is currently answering 503 (it prints "This
# endpoint is being retired"), so npm retries it: on 2026-09-04 that turned a
# 4s install into a 6m one and timed the CI job out mid-install. Measured on one
# machine, one lock file, one warm cache:
#
#   npm ci                                    61s
#   npm ci --no-audit --no-fund                4s
#   NPM_CONFIG_AUDIT=false NPM_CONFIG_FUND=false npm ci    4s
#
# Nothing is lost. `npm ci` exits 0 whatever the audit says (verified directly),
# the endpoint returns no advisories at all right now, and Dependabot already
# watches this same tree — it is what reported the 7 open npm alerts. This is CI
# only; a developer's `npm ci` still audits when the endpoint recovers.
env:
  NPM_CONFIG_AUDIT: "false"
  NPM_CONFIG_FUND: "false"

jobs:
  hash-lock:
    name: Hash lock / ${{ matrix.name }}
    runs-on: ${{ matrix.runner }}
    timeout-minutes: 10
    strategy:
      fail-fast: false
      matrix:
        include:
          - name: Linux x64
            runner: ubuntu-24.04
            architecture: x64
          - name: macOS ARM64
            runner: macos-15
            architecture: arm64
          - name: Windows x64
            runner: windows-2025
            architecture: x64

    steps:
      # actions/checkout v7.0.1
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1

      # actions/setup-python v7.0.0
      - uses: actions/setup-python@5fda3b95a4ea91299a34e894583c3862153e4b97
        with:
          python-version: "3.11"
          architecture: ${{ matrix.architecture }}
          cache: pip

      - name: Verify hash-locked dependencies
        # Both locks, separately: each must be internally complete on its own,
        # because the Docker build installs them as two independent
        # --require-hashes passes. A native matrix is required because package
        # markers are evaluated for the host running pip.
        run: |
          python -m pip install --dry-run --ignore-installed --require-hashes -r requirements-lock.txt
          python -m pip install --dry-run --ignore-installed --require-hashes -r requirements-channels-lock.txt

  test:
    name: test / Python ${{ matrix.python-version }}
    runs-on: ubuntu-latest
    # 10 minutes stopped fitting again after the frontend split below: on
    # 2026-09-29 the 3.11 leg's pytest step took 520s of a job that finished
    # at exactly 600s (run 36553260060), and the next push was cancelled at
    # 99% of the suite with no failing test (36556726032). The suite, not a
    # hang, fills the budget, and runner speed varies by minutes (the same
    # tree's 3.14 leg took 5m14s one run and 8m32s the other), so the limit
    # leaves room for that instead of being trimmed to the last good run.
    timeout-minutes: 20
    strategy:
      fail-fast: false
      matrix:
        # The floor and the ceiling of `requires-python` in pyproject.toml.
        # The ceiling leg exists because the <3.14 cap came off in 0bb1d870:
        # supporting a version nothing exercises is how the cap got stale in
        # the first place. Raise the upper leg whenever the floor or a new
        # CPython release moves.
        python-version: ["3.11", "3.14"]

    steps:
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1

      - uses: actions/setup-python@5fda3b95a4ea91299a34e894583c3862153e4b97 # v7.0.0
        with:
          python-version: ${{ matrix.python-version }}
          cache: pip

      - name: Install dependencies
        # `openbb` is included so agent/tests/test_openbb_bridge actually runs:
        # those tests importorskip on openbb_ai, so without the extra the whole
        # OpenBB Workspace bridge would report green while never being exercised.
        # `stats` is included for the same reason: the quantlib econometrics
        # tests skip without statsmodels/arch, and skipped is not tested.
        run: pip install -e ".[dev,openbb,stats,anthropic]"

      - name: Verify optional extras under test are importable
        # Fails loudly if an extra stops resolving, instead of letting its
        # suite silently degrade to skips.
        run: |
          python -c "import openbb_ai, importlib.metadata as m; print('openbb-ai', m.version('openbb-ai'))"
          python -c "import statsmodels; print('statsmodels', statsmodels.__version__)"
          python -c "import arch; print('arch', arch.__version__)"
          python -c "import langchain_anthropic; print('langchain-anthropic', langchain_anthropic.__version__)"

      - name: Repository safety gates
        run: bash tools/ci_grep_gates.sh

      - name: Test environment variable gate
        run: pytest tools/test_ci_env_var_gate.py -q

      - name: Syntax check
        run: |
          cd agent
          python -m compileall -q cli
          python -m py_compile api_server.py
          python -m py_compile mcp_server.py
          python -m py_compile src/agent/loop.py
          python -m py_compile src/tools/__init__.py
          python -m py_compile backtest/runner.py

      - name: Run tests
        # test_e2e_harness_v2.py is gated by VIBE_TRADING_RUN_LIVE_E2E
        # (real-LLM only); list --ignore here so CI's intent is explicit and
        # the suite never silently switches state on env presence.
        run: pytest --ignore=agent/tests/e2e_backtest --ignore=agent/tests/test_e2e_harness_v2.py --cov=agent --cov-report=term-missing --cov-report=xml --tb=short -q

  frontend:
    name: frontend
    runs-on: ubuntu-latest
    timeout-minutes: 10
    # Split out of `test` on 2026-09-04. It used to be the tail of BOTH Python
    # matrix legs, which ran the identical npm ci + tsc + vite + vitest twice
    # per push for a tree that has no Python in it. That duplication was free
    # until the Python 3.11 leg grew to 539s of its 600s budget: on 8555a05a an
    # npm registry slowdown pushed `npm ci` from its usual ~20s to over 130s and
    # the leg hit `timeout-minutes` mid-install, reporting as `cancelled` with a
    # green backend. Nothing here is tested less than before — the same three
    # commands run on the same tree — but the frontend now owns its own budget
    # and the Python legs get ~65s of headroom back.
    steps:
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1

      - uses: actions/setup-node@820762786026740c76f36085b0efc47a31fe5020 # v7.0.0
        with:
          node-version: "22"
          cache: npm
          cache-dependency-path: frontend/package-lock.json

      - name: Frontend build
        run: |
          cd frontend
          npm ci
          npm run build

      - name: Frontend tests
        run: |
          cd frontend
          npx vitest run --reporter=verbose

  windows-background:
    name: ${{ matrix.name }} / Python 3.14 background lifecycle
    runs-on: ${{ matrix.runner }}
    timeout-minutes: 10
    strategy:
      fail-fast: false
      matrix:
        include:
          - name: Windows 11 ARM64
            runner: windows-11-arm
            architecture: arm64
          - name: Windows Server 2025 x64
            runner: windows-2025
            architecture: x64

    steps:
      # actions/checkout v7.0.1
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1

      # actions/setup-python v7.0.0
      - uses: actions/setup-python@5fda3b95a4ea91299a34e894583c3862153e4b97
        with:
          python-version: "3.14"
          architecture: ${{ matrix.architecture }}
          cache: pip

      - name: Install focused test dependencies
        run: python -m pip install pytest pydantic pyyaml python-dotenv

      - name: Run Windows background safety regressions
        env:
          PYTHONPATH: agent
        run: >-
          python -m pytest
          agent/tests/test_background_tools.py
          agent/tests/test_bash_tool.py
          -q

  desktop-electron-windows:
    name: Windows desktop source lifecycle
    runs-on: windows-2025
    timeout-minutes: 30

    steps:
      # actions/checkout v7.0.1
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1

      # actions/setup-python v7.0.0
      - uses: actions/setup-python@5fda3b95a4ea91299a34e894583c3862153e4b97
        with:
          python-version: "3.12"
          architecture: x64
          cache: pip

      # actions/setup-node v7.0.0
      - uses: actions/setup-node@820762786026740c76f36085b0efc47a31fe5020
        with:
          node-version: "22"
          cache: npm
          cache-dependency-path: desktop/electron/package-lock.json

      - name: Install Vibe-Trading source backend
        run: python -m pip install -e .

      - name: Install desktop dependencies
        working-directory: desktop/electron
        run: npm ci

      - name: Prepare verified Electron runtime
        working-directory: desktop/electron
        run: npm run prepare:electron

      - name: Build desktop shell
        working-directory: desktop/electron
        run: npm run build

      - name: Run desktop lifecycle and parent-death smoke tests
        working-directory: desktop/electron
        run: npm run smoke:lifecycle