The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs users to run shell commands using the belt CLI, such as belt login, belt app run, and belt app store. These commands are standard operations for authenticating with the platform and executing image generation models.\n- [EXTERNAL_DOWNLOADS]: The skill references external configuration and installation guides from the inference-sh GitHub repository and the inference.sh website. These links are provided to assist the user with setting up the required command-line environment and are considered neutral references to the service's official documentation.\n- [PROMPT_INJECTION]: The skill processes user-provided prompts that are interpolated into shell commands. Although this represents an attack surface for indirect prompt injection, the risk is mitigated by the following factors:\n
Ingestion points: Prompt values inside the --input JSON argument in SKILL.md.\n
Boundary markers: User prompts are encapsulated within a structured JSON string literal, which helps maintain data separation.\n
Capability inventory: The skill uses the Bash tool to execute belt commands, which is the primary intended functionality.\n
Sanitization: The skill relies on the belt CLI to handle and sanitize the provided JSON input before processing it through generative models.

ai-image-generation