The Agent Skills Directory

[COMMAND_EXECUTION]: The bundled Node.js scripts in the scripts/ directory use child_process.spawnSync to execute the system utilities convert and identify (ImageMagick). These commands are used to perform image transformations such as background removal, resizing, and color quantization based on parameters defined in the skill's workflow.
[DYNAMIC_EXECUTION]: The file references/imagegen-cli-recovery.md provides Python scripts intended for the agent to execute to recover Base64-encoded image data from JSONL session logs. This is a maintenance utility for the platform and uses static templates provided within the skill.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user input via the subject parameter, which is interpolated into prompts for an image generation tool in references/prompt-patterns.md.
Ingestion points: User input subject in SKILL.md.
Boundary markers: Prompts use structured headers like Subject: and Style: to delimit the user input.
Capability inventory: The skill has the capability to execute shell commands via ImageMagick and write files to the local assets directory.
Sanitization: The skill relies on the structured prompt template and the image generator's internal safety filters.

generating-dot-assets