Omni Apps is introducing a new AI agent skill designed to streamline your development workflow and ensure the robustness of your projects. This skill, named testing, focuses on validating skill conformance and executing comprehensive test suites. It provides important health intelligence through regression-classified reports, helping teams maintain high quality standards across all their Omni Apps skills and associated projects. Developers and project managers can use this tool to gain deeper insights into their code's stability and consistency.\n\n

Comprehensive Validation and Analysis Modes\n\nThe testing skill operates in two distinct and complementary modes to cover different aspects of quality assurance. The first mode focuses entirely on skill conformance. It validates that every skill within your project has a proper documentation structure, includes necessary frontmatter, and achieves full manifest coverage. This initial validation step is fundamental for maintaining consistency, discoverability, and adherence to internal standards across your entire skill base, catching structural issues before they impact functionality or documentation quality.\n\nThe second mode is dedicated to comprehensive test execution and intelligent analysis. In this mode, it runs a project's tiered test suite, encompassing unit tests, evaluations, and broader system health checks. This ensures a thorough examination of your application's behavior at various levels. Critically, after these test runs, the tool doesn't just report failures; it analyzes git history. This historical analysis allows it to classify any identified failures accurately. Issues are categorized as regressions, indicating a new or re-introduced bug; stale tests, suggesting outdated test cases; flakes, highlighting intermittent or unreliable tests; or infrastructure issues, pointing to environmental problems. This detailed classification provides a clear and actionable picture of where development attention is most needed, moving beyond simple pass/fail reporting.\n\n

Practical Interaction and Triggering\n\nInteracting with the testing skill is designed to be straightforward and integrate reliably into daily development practices. You can activate it using a set of clear, natural language triggers. For instance, if you need to perform an initial check on the structural integrity of your Omni Apps skills, you might simply use the prompt 'validate skills'. To initiate a complete execution of your project's various test suites, 'test skills' will start the process. For a quick status update on ongoing runs or recent results, 'how are the tests' provides a summary. If you need to identify specific problems, 'what's broken' drills down into current issues and their classifications. Additionally, for routine quality checks, you can use 'daily test run' to keep a consistent pulse on project quality and catch issues early.\n\nConsider a practical scenario: your team is working on a complex feature, and multiple developers are contributing. Before merging, you could trigger 'test skills' to run the full suite. If the tool reports failures, its ability to analyze git history and classify them as, for example, a regression, helps immediately direct the developer to investigate recent changes. This intelligence saves time compared to manually sifting through logs. Another example: a developer asks 'what's broken' after a major refactor. The skill might report several 'stale tests' alongside a few 'regressions', allowing the developer to prioritize fixing the actual bugs before updating outdated test cases.\n\n

Intelligent Reporting and Enforced Quality\n\nThe testing skill goes beyond basic test execution by producing detailed reports enhanced with health intelligence. These reports are not just lists of passed or failed tests; they actively classify failures using the aforementioned categories: regressions, stale tests, flakes, and infrastructure issues. This precise classification empowers developers to understand the root cause of failures more quickly and prioritize their efforts effectively. For instance, a regression demands immediate attention, while a stale test might indicate a need for a test suite update. The skill also maintains state across successive runs, enabling robust trend analysis. This allows teams to observe patterns over time, track improvements in test suite health, and identify recurring issues that might require deeper architectural review.\n\nFurthermore, the platform actively enforces quality conventions across your projects. This means it helps ensure that specific standards and best practices for skill structure, documentation, and testing are met consistently. While the tool is powerful in its analytical capabilities, it respects the necessity for human judgment in complex or sensitive scenarios. For example, it is designed not to automatically fix security-test failures; such critical issues always require a human to review and address. Similarly, the skill will not un-skip tests without a clear understanding of the context, as making such changes automatically could inadvertently introduce new problems. Uncertain cases, particularly those involving security or significant test suite changes, are flagged for human review, ensuring that critical decisions remain with the development team.\n\n

Frequently Asked Questions\n\nQ: How does this skill classify test failures?\n\nA: The skill analyzes git history in conjunction with test results. It identifies whether a failure is a new issue (regression), an outdated test (stale test), an intermittent failure (flake), or an environmental problem (infrastructure issue).\n\nQ: Can the skill automatically fix my failing tests?\n\nA: No, the skill identifies issues and provides detailed classifications and reports, but it does not automatically fix code or test failures. Specifically, it does not auto-fix security-test failures or un-skip tests without understanding context, as these uncertain cases require human judgment.\n\nQ: What kind of reports can I expect from this skill?\n\nA: You can expect regression-classified reports with health intelligence. These reports highlight regressions, stale tests, flakes, and infrastructure issues. The skill also keeps state for trend analysis, allowing you to track test health over time.\n\nThis skill is designed to integrate into your workflow, providing clarity on test health and skill conformance without adding unnecessary complexity. It serves as a consistent partner in maintaining the quality and stability of your Omni Apps projects, allowing your team to focus on building features with confidence.