Run Parallel CI Checks Before an Agent Review

Use this recipe when lint and tests can run independently, but an agent review should start only after both checks pass.

Before you begin

You need:

• A project connected to GitHub. Find its integration connection slug in Settings → Integrations.
• An online runner with the ubuntu-latest label.
• A configured model provider and workspace agent defaults.
• A Node.js repository with lint and test scripts.

Add the workflow

Create .shipfox/workflows/parallel-ci-review.yml. Replace github_acme with your GitHub integration connection slug. This is a complete workflow:

# yaml-language-server: $schema=https://www.shipfox.io/docs/workflow.schema.json
name: Parallel CI and agent review
runner: ubuntu-latest

triggers:
  on_push:
    source: github_acme # Replace with your GitHub integration connection slug.
    event: push

jobs:
  lint:
    steps:
      - run: npm ci
      - run: npm run lint

  test:
    steps:
      - run: npm ci
      - run: npm test

  review:
    needs: [lint, test]
    steps:
      - prompt: Review the repository for likely bugs and missing test coverage.

Commit and push the file to the project's default branch. Wait for Parallel CI and agent review to sync in the project's Workflows tab.

Verify the job graph

Push another commit and open the new run:

1. Confirm that lint and test start without waiting for each other.
2. Confirm that review starts only after both checks pass.
3. Open the agent step and confirm that it returns a repository review.
4. Make one check fail and confirm that Shipfox skips review.

Each job receives a fresh checkout. The review job does not receive the lint or test logs. Use declared outputs for small values. Store reports and other artifacts outside the job checkout.

The extra checkout and dependency install in each job cost time. In return, independent checks can use separate runners and finish sooner. See Jobs and steps for the mental model and the workflow schema for exact dependency fields.