The Agent Skills Directory

[NO_CODE]: The skill consists entirely of instructional documentation in SKILL.md and does not contain any executable scripts, binaries, or configuration files.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied text prompts to generate or edit images through the model_run tool.
Ingestion points: User prompts are passed as parameters to image generation models (4.5, V3, etc.) as seen in the worked examples in SKILL.md.
Boundary markers: No explicit boundary markers or instruction-ignoring delimiters are provided for the prompt data.
Capability inventory: The skill uses tools including model_run, model_schema_get, asset_display, and asset_download to perform model operations and manage generated assets.
Sanitization: No specific sanitization or validation of the input prompt is described before it is interpolated into model parameters.

scenario-ideogram