The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to perform shell operations using the opencli command-line tool. User-supplied inputs, including prompts and local file paths, are interpolated directly into shell commands (e.g., opencli jimeng generate-image2image "<prompt>" --images /path/a.png). While this is the intended function of the skill, it creates a surface for command injection if inputs are not properly escaped by the agent.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data provided by users during the image generation workflow.
Ingestion points: User-provided image descriptions (prompts) and local file paths are ingested in SKILL.md to be passed to the opencli tool.
Boundary markers: The skill does not define specific delimiters or instructions to prevent the agent from interpreting instructions that might be embedded within user-supplied prompts.
Capability inventory: The skill possesses the capability to execute shell commands via opencli, which can perform network requests to image generation services and interact with the local file system.
Sanitization: No sanitization or validation logic is defined within the skill to filter potentially malicious sequences in the user prompt before execution.

dreamina-opencli-image2image