The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill guides the user to install several packages and tools, including npm packages (flutter-skill, flutter-skill-react-native, flutter-skill-electron), Homebrew formulae (ai-dashboad/flutter-skill), and SDKs from Maven/NuGet. These resources are official components provided by the author ai-dashboad.
[COMMAND_EXECUTION]: The skill involves the execution of various CLI commands such as flutter-skill init to patch projects and flutter-skill launch to start the testing bridge. It also requires configuring the AI agent to execute the flutter-skill server binary.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection because it ingests data from external running applications that the agent then acts upon. \n
Ingestion points: Application UI elements are retrieved via inspect_interactive(), application logs via get_logs(), and visual states via screenshot(). \n
Boundary markers: The skill does not provide instructions to the agent to treat data from the app (such as button labels, text field values, or log messages) as untrusted content that should not be followed as instructions. \n
Capability inventory: The agent has the ability to perform impactful actions like tap(), enter_text(), and press_key() based on its interpretation of the app's state. \n
Sanitization: There is no evidence of filtering or sanitization of the content extracted from the application environment.

e2e-testing