The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references standard, well-known libraries for mobile automation.
Mentions official ecosystem packages such as Appium-Python-Client, pytest, and @wdio/appium-service.
Includes Maven coordinates for standard Java testing libraries including io.appium:java-client and org.seleniumhq.selenium:selenium-java.
[COMMAND_EXECUTION]: The skill references execution of the Appium server and CLI tools for application management.
Documentation mentions the --relaxed-security flag for the Appium CLI, which is a standard feature for enabling advanced device interactions (like shell commands or screen recording) during testing.
Provides curl command examples for uploading application files to the LambdaTest cloud infrastructure.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided application details and test scenarios to generate automation scripts.
Ingestion points: User-provided descriptions of mobile applications and test flows are ingested via the agent prompt.
Boundary markers: No explicit boundary markers or "ignore embedded instructions" warnings are used when interpolating user requirements into the generated code templates.
Capability inventory: The skill generates code that can perform network operations (to the mobile hub), file system interactions on the mobile device (push/pull), and executes shell-like interactions via the Appium driver.
Sanitization: The skill relies on the underlying LLM's inherent safety training for sanitization of user-provided strings during code generation.
[SAFE]: Cloud integration and network operations target the vendor's official infrastructure.
Operations involving lambdatest.com and its subdomains (e.g., mobile-hub.lambdatest.com, manual-api.lambdatest.com) are legitimate interactions for a skill authored by 'lambdatest'.

appium-skill