Overview
An end-to-end automation pipeline that detects test failures, diagnoses the problem, fixes the code, and opens a merge request—all without human intervention. Built in 3 days during an Indeed hackathon.
When a Playwright test fails, the system captures the full stack trace and context, then hands it to an AI agent. The agent analyzes the failure, generates a fix, commits it to a new branch, opens a GitLab merge request, creates a Jira ticket with debugging details, and comments on the ticket when the MR is ready for review.
The result: developers wake up to a fix already waiting for approval instead of a red build and a morning of debugging.
Gallery
Image 1 Placeholder
Image 2 Placeholder
Image 3 Placeholder
Key Features
- AI-powered failure analysis: parses stack traces and identifies root cause
- Automatic code fixes generated and validated before commit
- Full GitLab integration: branch creation, commits, and merge requests
- Jira ticket creation with stack traces, screenshots, and debugging context
- Automated MR comments notifying when fix is ready for review
- Multiple trigger modes: CLI, webhook, or form-based execution
- Cross-browser testing with Chromium, Firefox, and WebKit
- Video recording and trace capture for every test run
Highlights
3 days
Built In
2025
Hackathon
End-to-end
Automation
Technologies
n8nPlaywrightOpenAI APITypeScriptGitLab APIJira APIWebhooksDocker