Playwright Test Healer

Overview

This agent acts as a specialized Playwright test healer for the Opik application. Its primary function is to systematically diagnose and resolve failing End-to-End (E2E) tests, ensuring that fixes adhere strictly to established coding conventions.

The workflow involves running identified failures, debugging the state using browser snapshots, determining if the root cause lies in the test code or the application logic, applying necessary fixes, and verifying success through iterative re-runs.

Capabilities

• Test Execution & Debugging: Uses dedicated tools to run and debug specific failing tests within the Playwright framework.
• State Investigation: Leverages browser snapshotting capabilities to accurately understand the UI state at the point of failure.
• Issue Diagnosis: Identifies common Opik-specific issues, such as eventual consistency timing problems or environment setup failures (e.g., Flask helper service).
• Code Correction: Edits test code directly to implement necessary waits, retries, or structural fixes while respecting existing Page Object Model patterns.

Example Use Cases

1. Timing Failures: If a test fails because the UI hasn't updated after an asynchronous SDK call (e.g., creating a project), this agent will inject appropriate explicit waits (await helperClient.waitForProjectVisible(...)) to ensure data visibility before proceeding.
2. Environment Flakiness: When connection errors occur due to missing backend services, it can correctly identify these as environment issues and mark the test appropriately (test.fixme()).
3. Complex Workflow Fixes: For navigation or state transitions that fail intermittently, it will trace the path through the Page Object Catalog to apply robust fixes, ensuring all conventions are met.