The Agent Skills Directory

[SAFE]: The skill uses the astria CLI toolset, which corresponds to the vendor's (astriaai) own infrastructure for image generation services. All commands such as astria prompts, astria tunes, and astria generate are consistent with the skill's stated purpose.
[COMMAND_EXECUTION]: The skill utilizes Bash commands to interact with the Astria API. These operations are scoped to the vendor-provided astria utility and do not involve arbitrary shell execution or unauthorized system access.
[INDIRECT_PROMPT_INJECTION]: The skill constructs CLI commands using user-provided descriptions for image generation.
Ingestion points: User input regarding image content, subjects, and styling preferences (SKILL.md).
Boundary markers: The instructions explicitly require the agent to "show the user the prompt text and let them review or edit it before generating," ensuring user oversight.
Capability inventory: Employs the astria CLI for listing resources, creating tunes, and triggering generations.
Sanitization: Implements a human-in-the-loop review process where the final generated prompt and parameters must be confirmed by the user before execution.
[SAFE]: No obfuscation, hardcoded credentials, persistence mechanisms, or unauthorized remote code execution patterns were identified in the skill.

prompt-writing