ci.yml

name: CI

on:
  push:
    branches: [main, claude/test-aws-gpu-ci-gl5Jf]
  pull_request:
    branches: [main]

concurrency:
  group: gpu-${{ github.head_ref || github.ref }}
  cancel-in-progress: true

permissions:
  contents: read

jobs:
  lint:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: "3.12"
      - name: Install uv
        uses: astral-sh/setup-uv@v4
      - name: Install dependencies
        run: uv pip install --system ruff mypy
      - name: Ruff lint
        run: ruff check .
      - name: Ruff format check
        run: ruff format --check .
      - name: Mypy type check
        run: mypy src/

  test:
    runs-on: ubuntu-latest
    strategy:
      matrix:
        python-version: ["3.10", "3.11", "3.12", "3.13"]
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: ${{ matrix.python-version }}
      - name: Install uv
        uses: astral-sh/setup-uv@v4
      - name: Install package with dev dependencies
        run: uv pip install --system -e ".[dev]"
      - name: Run tests with coverage
        run: pytest

  test-mujoco:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: "3.12"
      - name: Install uv
        uses: astral-sh/setup-uv@v4
      - name: Install system dependencies (OSMesa for headless rendering)
        run: sudo apt-get update && sudo apt-get install -y libosmesa6-dev
      - name: Install package with demo and dev dependencies
        run: uv pip install --system -e ".[demo,dev]"
      - name: Run tests (including MuJoCo-dependent tests)
        env:
          MUJOCO_GL: osmesa
        run: pytest

  mujoco-example:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: "3.12"
      - name: Install uv
        uses: astral-sh/setup-uv@v4

      - name: Install system dependencies (OSMesa for headless rendering)
        run: sudo apt-get update && sudo apt-get install -y libosmesa6-dev

      - name: Install package with demo dependencies
        run: uv pip install --system -e ".[demo]"

      - name: Run MuJoCo grasp example
        env:
          MUJOCO_GL: osmesa
        run: python examples/demos/mujoco/grasp.py --report --assert-success --output-dir ./harness_output

      - name: Upload visual artifacts
        if: always()
        uses: actions/upload-artifact@v4
        with:
          name: mujoco-grasp-output
          path: |
            harness_output/report.html
            harness_output/mujoco_grasp/trial_001/autonomous_report.json
            harness_output/mujoco_grasp/trial_001/alarms.json
            harness_output/mujoco_grasp/trial_001/phase_manifest.json
            harness_output/mujoco_grasp/trial_001/**/*_rgb.png
            harness_output/mujoco_grasp/trial_001/**/metadata.json
          retention-days: 30

      - name: Post summary with checkpoint images
        if: always()
        run: |
          echo "## MuJoCo Grasp Example Results" >> $GITHUB_STEP_SUMMARY
          echo "" >> $GITHUB_STEP_SUMMARY
          echo "Checkpoint captures from the scripted grasp sequence:" >> $GITHUB_STEP_SUMMARY
          echo "" >> $GITHUB_STEP_SUMMARY
          for cp in harness_output/mujoco_grasp/trial_001/*/; do
            name=$(basename "$cp")
            echo "### ${name}" >> $GITHUB_STEP_SUMMARY
            echo "" >> $GITHUB_STEP_SUMMARY
            if [ -f "${cp}front_rgb.png" ]; then
              echo "![${name} front view](data:image/png;base64,$(base64 -w0 "${cp}front_rgb.png"))" >> $GITHUB_STEP_SUMMARY
            fi
            if [ -f "${cp}metadata.json" ]; then
              step=$(python3 -c "import json; print(json.load(open('${cp}metadata.json'))['step'])")
              sim_time=$(python3 -c "import json; print(f\"{json.load(open('${cp}metadata.json'))['sim_time']:.3f}\")")
              echo "" >> $GITHUB_STEP_SUMMARY
              echo "Step: ${step} | Sim time: ${sim_time}s" >> $GITHUB_STEP_SUMMARY
            fi
            echo "" >> $GITHUB_STEP_SUMMARY
          done
          echo "---" >> $GITHUB_STEP_SUMMARY
          echo "Download the full artifact (including HTML report) from the Actions tab." >> $GITHUB_STEP_SUMMARY

  lerobot-g1-native-example:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: "3.12"
      - name: Install uv
        uses: astral-sh/setup-uv@v4

      - name: Install system dependencies (OSMesa for headless rendering)
        run: sudo apt-get update && sudo apt-get install -y libosmesa6-dev

      - name: Install package with demo dependencies
        run: uv pip install --system -e ".[demo]"

      - name: Install unitree SDK directly from fork
        run: |
          # unitree-sdk2py is installed directly (not via a pyproject extra —
          # PyPI rejects packages with git-based direct references on upload)
          uv pip install --system "unitree-sdk2py @ git+https://github.com/MiaoDX-fork-and-pruning/unitree_sdk2_python_uv.git"

      - name: Install CPU PyTorch and lerobot
        run: |
          uv pip install --system torch --index-url https://download.pytorch.org/whl/cpu
          uv pip install --system lerobot

      - name: Install lerobot/unitree-g1-mujoco runtime deps
        run: uv pip install --system loguru pygame scipy pyyaml

      - name: Cache HuggingFace models
        uses: actions/cache@v4
        with:
          path: ~/.cache/huggingface/hub
          key: hf-models-${{ runner.os }}-v1

      - name: Run native LeRobot G1 example (GR00T)
        env:
          MUJOCO_GL: osmesa
        run: python examples/demos/g1/lerobot_native.py --controller groot --report --assert-success --output-dir ./harness_output

      - name: Upload visual artifacts
        if: always()
        uses: actions/upload-artifact@v4
        with:
          name: lerobot-g1-native-output
          path: |
            harness_output/lerobot_g1_native_groot_report.html
            harness_output/lerobot_g1_native_groot/trial_001/**/*_rgb.png
            harness_output/lerobot_g1_native_groot/trial_001/**/metadata.json
          retention-days: 30

  gpu-changes:
    runs-on: ubuntu-latest
    outputs:
      gpu_relevant: ${{ steps.filter.outputs.gpu }}
    steps:
      - uses: actions/checkout@v4
      - uses: dorny/paths-filter@v3
        id: filter
        with:
          filters: |
            gpu:
              - 'tests/integration/gpu/test_gpu_smoke.py'
              - 'tests/integration/gpu/test_isaac_lab_gpu.py'
              - 'tests/unit/wrappers/test_isaac_lab_compat.py'
              - 'src/roboharness/wrappers/gymnasium_wrapper.py'
              - '.cirun.yml'
              - 'Dockerfile.gpu'
              - '.github/workflows/ci.yml'

  # ─── gpu-test (GCP) PAUSED 2026-04-14 ────────────────────────────────
  # GCP runner is currently paused in .cirun.yml; AWS Sydney is the
  # primary GPU CI. To re-enable, uncomment the block below AND
  # uncomment the GCP runner block in .cirun.yml.
  #
  # gpu-test:
  #   needs: [gpu-changes]
  #   runs-on: [self-hosted, cirun-gpu]
  #   timeout-minutes: 45
  #   # Only one GCP GPU job runs at a time globally — we only have quota for
  #   # a single T4. `cancel-in-progress: true` so a new push immediately
  #   # pre-empts any stale in-progress / queued job in this group; otherwise
  #   # a stuck provisioning attempt would hold the slot indefinitely and
  #   # every subsequent push would pile up behind it.
  #   concurrency:
  #     group: gpu-test-gcp
  #     cancel-in-progress: true
  #   if: |
  #     github.event_name == 'push' && github.ref == 'refs/heads/main'
  #     || contains(github.event.pull_request.labels.*.name, 'gpu-test')
  #     || needs.gpu-changes.outputs.gpu_relevant == 'true'
  #   steps:
  #     - uses: actions/checkout@v4
  #     - uses: actions/setup-python@v5
  #       with:
  #         python-version: "3.12"
  #     - name: Install uv
  #       uses: astral-sh/setup-uv@v4
  #     - name: Install package with dev dependencies and torch
  #       run: |
  #         uv pip install --system -e ".[dev]"
  #         uv pip install --system torch --index-url https://download.pytorch.org/whl/cu121
  #     - name: Verify GPU
  #       run: |
  #         nvidia-smi
  #         python -c 'import torch; print(f"CUDA: {torch.cuda.is_available()}, GPUs: {torch.cuda.device_count()}")'
  #     - name: Run GPU tests
  #       run: pytest -m gpu -v --no-cov
  # ─── End paused gpu-test (GCP) ───────────────────────────────────────

  # AWS GPU smoke test — primary GPU CI. Runs on the cirun-aws-gpu runner
  # provisioned on AWS (g4dn.xlarge in ap-southeast-2 + NVIDIA GPU-Optimized
  # AMI). Gated to the test branch and to the gpu-relevant paths filter so
  # we don't spin up AWS instances on unrelated pushes.
  gpu-test-aws:
    needs: [gpu-changes]
    runs-on: [self-hosted, cirun-aws-gpu]
    timeout-minutes: 45
    # Serialize AWS GPU jobs too. `cancel-in-progress: true` so a new push
    # pre-empts any stale in-progress / queued attempt — important because
    # Cirun provisioning on AWS can silently sit queued (e.g. if the
    # Dashboard has no AWS credentials or the Marketplace AMI subscription
    # is still propagating), and `false` would stack pushes behind forever.
    # Runs in a different group from gpu-test-gcp, so GCP and AWS do not
    # block each other.
    concurrency:
      group: gpu-test-aws
      cancel-in-progress: true
    if: |
      github.ref == 'refs/heads/claude/test-aws-gpu-ci-gl5Jf'
      || contains(github.event.pull_request.labels.*.name, 'gpu-test-aws')
      || needs.gpu-changes.outputs.gpu_relevant == 'true'
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: "3.12"
      - name: Install uv
        uses: astral-sh/setup-uv@v4
      - name: Show host info
        run: |
          uname -a
          cat /etc/os-release || true
          nvidia-smi
      - name: Install package with dev dependencies and torch
        run: |
          uv pip install --system -e ".[dev]"
          uv pip install --system torch --index-url https://download.pytorch.org/whl/cu121
      - name: Verify PyTorch CUDA
        run: |
          python -c 'import torch; print(f"torch={torch.__version__}, CUDA available={torch.cuda.is_available()}, GPUs={torch.cuda.device_count()}, name={torch.cuda.get_device_name(0) if torch.cuda.is_available() else None}")'
      - name: Run GPU tests
        run: pytest -m gpu -v --no-cov

  lerobot-g1-example:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: "3.12"
      - name: Install uv
        uses: astral-sh/setup-uv@v4

      - name: Install system dependencies (OSMesa for headless rendering)
        run: sudo apt-get update && sudo apt-get install -y libosmesa6-dev

      - name: Install package with demo dependencies
        run: uv pip install --system -e ".[demo]"

      - name: Run LeRobot G1 example
        env:
          MUJOCO_GL: osmesa
        run: python examples/demos/g1/lerobot_locomotion.py --report --assert-success --output-dir ./harness_output

      - name: Upload visual artifacts
        if: always()
        uses: actions/upload-artifact@v4
        with:
          name: lerobot-g1-output
          path: |
            harness_output/lerobot_g1_report.html
            harness_output/lerobot_g1/trial_001/**/*_rgb.png
            harness_output/lerobot_g1/trial_001/**/metadata.json
          retention-days: 30

      - name: Post summary
        if: always()
        run: |
          echo "## LeRobot G1 Example Results" >> $GITHUB_STEP_SUMMARY
          echo "" >> $GITHUB_STEP_SUMMARY
          echo "Checkpoint captures from the G1 humanoid validation:" >> $GITHUB_STEP_SUMMARY
          echo "" >> $GITHUB_STEP_SUMMARY
          for cp in harness_output/lerobot_g1/trial_001/*/; do
            name=$(basename "$cp")
            n_images=$(find "$cp" -name '*_rgb.png' 2>/dev/null | wc -l)
            if [ -f "${cp}metadata.json" ]; then
              step=$(python3 -c "import json; print(json.load(open('${cp}metadata.json'))['step'])")
              echo "- **${name}**: step ${step}, ${n_images} captures" >> $GITHUB_STEP_SUMMARY
            else
              echo "- **${name}**: ${n_images} captures" >> $GITHUB_STEP_SUMMARY
            fi
          done
          echo "" >> $GITHUB_STEP_SUMMARY
          echo "Download the full artifact (including HTML report and images) from the Actions tab." >> $GITHUB_STEP_SUMMARY