integration-test
Full SDK integration test that runs actual queries through the Claude SDK sandbox. Use after making changes to SDK client code, session management, skill loading, network proxy, voice/TTS, or image generation. Runs real prompts through the SDK to verify the complete path works.
Security Assessment
About integration-test
The integration-test skill is designed to perform comprehensive end-to-end verification of the Claude SDK environment. It automates the process of running real queries through the SDK sandbox, allowing developers to confirm that the entire system—ranging from SDK client interactions to session management, skill loading, network proxies, voice/TTS processing, and image generation—functions correctly. By executing actual prompts, this skill ensures that changes to the SDK or related infrastructure do not break the expected workflows, providing a robust safety net for maintaining the stability and reliability of the system.
Key features of the integration-test skill include specialized tests for various components of the SDK environment. The Echo Test checks basic SDK communication, while the Environment Test verifies that environment variables like OPENAI_API_KEY are correctly passed into the sandbox. Network Test ensures that HTTP proxies are properly configured and capable of reaching external services. The Image Test confirms end-to-end image generation functionality, and the Voice Test validates compliance with voice message formatting and output rules. Finally, the Agent Transport Test verifies direct agent communication, covering session handling, RPC authentication, and transport mechanisms. Each test can be run individually or collectively, with options for verbose output and custom timeouts, allowing for detailed inspection and debugging.
This skill is primarily intended for developers and testers working on Claude SDK integrations or extensions. It is especially useful after making changes to SDK client code, session management, skill loading, or service modules such as voice/TTS and image generation. By automating critical validation steps, it saves time and reduces errors compared to manual testing. It is also valuable for maintaining consistent behavior across development, staging, and production environments, ensuring that all components function seamlessly before deployment or user-facing release.
FAQ
When should I run the integration-test skill?
You should run it after making changes to the SDK client code, session management, skill files, network proxy settings, voice/TTS modules, image generation services, or Telegram media handling.
Can I run individual tests instead of all tests?
Yes, you can run specific tests using flags such as --echo, --env, --network, --image, --voice, or --agents to focus on particular components.
Are there any environment variables required?
Some tests, like Image and Voice, require OPENAI_API_KEY. The Agent Transport test requires TELCLAUDE_AGENT_URL. The Environment Test checks for these variables or falls back to HOME if not set.
What does a Voice Test failure usually indicate?
It often means the audio was generated incorrectly, such as wrong format (.mp3 instead of .ogg), wrong directory, missing --voice-message flag, or excessive text output. Check that skills are correctly loaded and no conflicts exist.
Is verbose output available for debugging?
Yes, you can add the -v flag when running tests to see detailed execution logs and timing for each test.
Install integration-test
Quick Setup:
- Copy the skill folder to
.claude/skills/ - Claude will automatically detect and use the skill
Repository
avivsinai/telclaude