Integration Tests

What it does

This skill enables developers to run automated integration tests specifically designed for OpenAI Agents applications built with the openai-agents-js library. It validates that different components of an AI application interact correctly in real-world scenarios, ensuring reliability before deployment.

When to use it

• Before pushing code changes to a production environment to catch regressions early.
• During the development phase to verify that new agent logic integrates seamlessly with existing tools or workflows.
• When refactoring core agent architecture to ensure external dependencies and API calls remain functional.
• As part of a continuous integration (CI) pipeline to automatically gate releases based on test results.

Key capabilities

• Executes comprehensive integration tests tailored for the openai-agents-js framework.
• Validates end-to-end interactions between AI agents and their configured tools or APIs.
• Provides feedback on application stability and component connectivity within complex agent systems.

Example prompts

• "Run the full suite of integration tests for my current OpenAI Agents project to check for any broken workflows."
• "Execute integration tests specifically for the new tool I just added to verify its compatibility with the main agent."
• "Generate a report on integration test failures after I updated the agent's system instructions."

Tips & gotchas

Ensure your local environment matches the production configuration (e.g., API keys, database connections) before running these tests to avoid false positives. This skill is specific to the openai-agents-js library and may not be applicable to agents built with other frameworks or raw API calls.