The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes the dreamina CLI and various subcommands like text2image, session, and login to perform generation tasks and manage OAuth states. It also references the use of Maven (mvn test) in the contract verification documentation to ensure environment consistency.
[EXTERNAL_DOWNLOADS]: Documentation in the contract reference file points to the official Dreamina website (jimeng.jianying.com) for downloading and installing the CLI tool.
[REMOTE_CODE_EXECUTION]: The skill's reference documentation provides an installation command that pipes a remote script into a shell (curl | bash). While this is sourced from the official product domain, it represents a high-privilege execution pattern.
[PROMPT_INJECTION]: The skill processes untrusted user-provided text prompts for image and video generation, which creates a surface for indirect prompt injection.
Ingestion points: User-supplied prompts and local image/media file paths are used as arguments for CLI commands (SKILL.md).
Boundary markers: None identified in the instructions to separate user content from system instructions.
Capability inventory: The skill can execute various CLI subcommands via shell with arguments interpolated from user input.
Sanitization: No explicit sanitization or escaping mechanisms for user-provided input strings are documented.

dreamina-cli