Skip to content

TESTING-GUIDELINE.md

Latest commit

 

History

History
42 lines (33 loc) · 1.55 KB

TESTING-GUIDELINE.md

File metadata and controls

42 lines (33 loc) · 1.55 KB
Raw

Testing Guideline

This document describes the testing practices for Jakarta Agentic AI, including compatibility requirements and TCK usage.

Technology Compatibility Kit (TCK)

  • The TCK module verifies compatibility of implementations with the Jakarta Agentic AI specification.
  • All implementations must pass the TCK to be considered compliant.

Running Tests

  • Build and run the TCK with required upstream modules: 
    mvn --projects tck --also-make verify
  • If you need a full clean build of the TCK artifacts as well: 
    mvn --projects tck --also-make clean install
  • Add additional tests as needed to cover new features and edge cases.

Signaling Implementation Presence

The TCK contains two kinds of assertions:

  • Assertions tagged with @RequiresImplementation exercise behavior that only a compatible Jakarta Agentic AI implementation can dispatch (@Decision, @Action, @Outcome, @HandleException).
  • Assertions tagged with @RequiresNoImplementation verify the plain-CDI baseline (@Trigger observation only) and must be skipped when a compatible implementation is present.

The TCK detects implementation presence via a system property. Implementations running the TCK must set:

-Djakarta.ai.agent.tck.implementation.present=true

on the Surefire / Failsafe argLine (or globally on the test JVM). Leaving the property unset — the default — signals a plain-CDI run and skips the @RequiresImplementation assertions.

Reporting Issues

  • Please report any test failures or compatibility issues via GitHub Issues.