Skip to content

eval

eval #18

Workflow file for this run

name: eval
on:
pull_request:
branches: [main]
schedule:
# Nightly at 03:00 UTC against the default branch.
- cron: "0 3 * * *"
workflow_dispatch:
permissions:
contents: read
pull-requests: write
jobs:
eval:
runs-on: ubuntu-latest
env:
# Replay mode never calls Gemini; a throwaway value keeps parity with CI.
GEMINI_API_KEY: ci-test-key
steps:
- uses: actions/checkout@v4
- uses: actions/setup-python@v5
with:
python-version: "3.11"
- name: Install dependencies
run: |
python -m pip install --upgrade pip
pip install -r requirements.txt
- name: Build golden dataset
run: python eval/golden/generate_golden.py
- name: Run evaluation (offline replay)
id: eval
run: python -m eval.run_eval
- name: Upload HTML report
if: always()
uses: actions/upload-artifact@v4
with:
name: eval-report
path: eval/report/latest/
if-no-files-found: warn
- name: Post PR summary comment
if: always() && github.event_name == 'pull_request'
env:
GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
PR: ${{ github.event.pull_request.number }}
run: |
if [ -f eval/report/latest/summary.md ]; then
gh pr comment "$PR" --body-file eval/report/latest/summary.md --edit-last \
|| gh pr comment "$PR" --body-file eval/report/latest/summary.md
fi