trunk-ci-setup

name	trunk-ci-setup
description	Best practices for configuring CI pipelines to upload test results to Trunk. Use when editing CI configuration files, setting up test reporting, or configuring JUnit XML output.

Trunk CI Setup Best Practices

When helping users configure CI pipelines or test reporting, follow these guidelines to ensure test results upload correctly to Trunk.

JUnit XML Output

Trunk ingests test results via JUnit XML format. Most test frameworks support this natively or via a plugin:

• Jest: Use jest-junit reporter → npm install --save-dev jest-junit, set JEST_JUNIT_OUTPUT_DIR
• Vitest: Built-in → reporters: ['junit'] in vitest.config.ts, set outputFile
• pytest: Built-in → --junitxml=report.xml flag
• Go: Use gotestsum → gotestsum --junitfile report.xml
• JUnit/Gradle: Built-in → results in build/test-results/test/
• JUnit/Maven: Built-in → results in target/surefire-reports/
• RSpec: Use rspec_junit_formatter gem
• PHPUnit: Built-in → --log-junit report.xml flag

Always configure the output path explicitly rather than relying on defaults — this makes the upload step reliable.

Trunk Analytics Uploader

After tests run, upload results using the trunk-analytics-cli:

# Install (one-time, add to CI)
curl -fsSL https://trunk.io/releases/analytics-cli/latest/trunk-analytics-cli-linux-x86_64.tar.gz | tar xz

# Upload
trunk-analytics-cli upload \
  --org-url-slug <your-org> \
  --token $TRUNK_TOKEN \
  --junit-paths "path/to/reports/**/*.xml"

The TRUNK_TOKEN should be stored as a CI secret, never hardcoded.

CI Provider Tips

GitHub Actions

• Use if: always() on the upload step so results are uploaded even when tests fail — this is critical for capturing flaky test data.
• The test and upload steps should be in the same job to share the filesystem.

CircleCI

• Use when: always on the upload step.
• Store test results with store_test_results as well for CircleCI's built-in reporting.

GitLab CI

• Add the upload to an after_script block or use when: always on the upload job.
• Use artifacts:reports:junit for GitLab's native reporting alongside Trunk uploads.

Jenkins

• Add the upload as a post { always { } } block in the pipeline.
• Use the JUnit plugin for Jenkins-native reporting alongside Trunk.

Common Mistakes

• Not uploading on test failure: If the upload step only runs when tests pass, you'll never capture flaky test data. Always use if: always() / when: always.
• Wrong glob pattern: Double-check the JUnit XML path matches what the test framework actually produces.
• Missing token: The TRUNK_TOKEN must be set as a CI secret. Check the Trunk dashboard at https://app.trunk.io for your org's token.
• Multiple XML files: Some frameworks produce one XML per test suite. Use glob patterns (**/*.xml) to capture all of them.

Documentation

• Full setup guide: https://docs.trunk.io/flaky-tests/use-mcp-server/mcp-tool-reference/set-up-test-uploads
• Flaky Tests overview: https://docs.trunk.io/flaky-tests
• CI configuration examples: https://docs.trunk.io/flaky-tests