The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies extensively on executing the asc CLI tool to perform its core functions, such as listing templates, applying themes, and generating screenshot previews. These commands involve reading local screenshot files and writing output to HTML or PNG files.
[INDIRECT_PROMPT_INJECTION]: The skill accepts untrusted user input via command-line flags such as --headline, --subtitle, --tagline, and --prompt. This input is interpolated into prompts sent to the Gemini AI for enhancing screenshots or generating theme designs.
Ingestion points: User-supplied text flags in SKILL.md and references/commands.md (e.g., --headline, --prompt).
Boundary markers: The instructions do not specify any delimiters or safety warnings to prevent the AI from following malicious instructions embedded within the user-provided text.
Capability inventory: The skill has the capability to write files to the local system (--image-output) and perform network operations via API calls to Gemini.
Sanitization: There is no evidence of sanitization or validation of the input strings before they are processed by the AI.
[DATA_EXPOSURE]: The skill accesses a local configuration file located at ~/.asc/app-shots-config.json to manage Gemini API keys. While this is a common practice for CLI tools to handle secrets, it involves accessing a specific sensitive directory on the user's filesystem.

asc-app-shots