You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
 
 
 

44 lines
1.1 KiB

name: Agent Eval Check
on:
pull_request:
branches: [main, feature/agent-mvp]
paths:
- 'apps/api/src/app/agent/**'
- '.github/workflows/agent-eval.yml'
push:
branches: [main, feature/agent-mvp]
paths:
- 'apps/api/src/app/agent/**'
- '.github/workflows/agent-eval.yml'
jobs:
eval-check:
name: Deterministic Agent Eval
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Setup Node.js
uses: actions/setup-node@v4
with:
node-version: 22
cache: 'npm'
- name: Install dependencies
run: npm ci
- name: Check for snapshot file
id: check-snapshot
run: |
if [ -f "apps/api/src/app/agent/eval-snapshots.json" ]; then
echo "exists=true" >> $GITHUB_OUTPUT
else
echo "exists=false" >> $GITHUB_OUTPUT
echo "::warning::No eval-snapshots.json found. Run 'AGENT_EVAL_TOKEN=<jwt> npm run eval:snapshot' to generate."
fi
- name: Run deterministic eval checks
if: steps.check-snapshot.outputs.exists == 'true'
run: npm run eval:check