CI/CD Integration

Add LaunchGate eval gating to your continuous integration pipeline so quality checks run automatically on every change.

GitHub Actions (recommended)

The simplest integration uses the official GitHub Action.

Add your API key as a secret

Go to your repository Settings → Secrets and variables → Actions and add LAUNCHGATE_API_KEY.

Create the workflow

.github/workflows/eval.yml


name: LaunchGate Eval
 
on:
  pull_request:
    branches: [main]
 
permissions:
  pull-requests: write
 
jobs:
  eval:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
 
      - name: Set up Node.js
        uses: actions/setup-node@v4
        with:
          node-version: "22"
 
      - name: Install dependencies
        run: npm ci
 
      - name: Generate AI output
        id: generate
        run: |
          OUTPUT=$(node scripts/generate.js)
          echo "text=$OUTPUT" >> $GITHUB_OUTPUT
 
      - name: Run LaunchGate Eval
        uses: launchgate/eval-action@v1
        with:
          api-key: ${{ secrets.LAUNCHGATE_API_KEY }}
          suite: "rag-faithfulness"
          output: ${{ steps.generate.outputs.text }}
          input: '{"context": "...", "query": "..."}'

The action will:

Run the eval suite against your output
Post a formatted results comment on the PR
Fail the workflow if the eval is “aborted”

Using the CLI

For CI systems without a dedicated action, use the CLI:

GitHub Actions

.github/workflows/eval.yml


- name: Install LaunchGate CLI
  run: npm install -g @launchgate/cli
 
- name: Run eval
  env:
    LAUNCHGATE_API_KEY: ${{ secrets.LAUNCHGATE_API_KEY }}
  run: |
    launchgate run rag-faithfulness \
      --output "${{ steps.generate.outputs.text }}" \
      --input '{"context": "..."}'

GitLab CI

.gitlab-ci.yml


eval:
  stage: test
  image: node:22
  variables:
    LAUNCHGATE_API_KEY: $LAUNCHGATE_API_KEY
  script:
    - npm install -g @launchgate/cli
    - |
      launchgate run rag-faithfulness \
        --output "$(node scripts/generate.js)" \
        --input '{"context": "..."}'

Generic


#!/bin/bash
# Works in any CI system with Node.js
 
export LAUNCHGATE_API_KEY="your-key-here"
 
npm install -g @launchgate/cli
 
OUTPUT=$(node scripts/generate.js)
 
launchgate run rag-faithfulness \
  --output "$OUTPUT" \
  --input '{"context": "..."}'
 
# Exit code 0 = cleared, 1 = aborted

Multiple suites

Run multiple eval suites in parallel:

.github/workflows/eval.yml


jobs:
  eval:
    runs-on: ubuntu-latest
    strategy:
      matrix:
        suite: [rag-faithfulness, content-safety, format-compliance]
    steps:
      - name: Run eval
        uses: launchgate/eval-action@v1
        with:
          api-key: ${{ secrets.LAUNCHGATE_API_KEY }}
          suite: ${{ matrix.suite }}
          output: ${{ needs.generate.outputs.text }}

Non-blocking mode

If you want eval results without blocking the pipeline:


- name: Run eval (non-blocking)
  uses: launchgate/eval-action@v1
  with:
    api-key: ${{ secrets.LAUNCHGATE_API_KEY }}
    suite: "my-suite"
    output: "..."
    fail-on-abort: "false"

Non-blocking mode still posts a PR comment with results — it just doesn’t fail the workflow step. This is useful during initial rollout when you want to observe eval results before enforcing them.