The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill downloads generated image content from provider CDNs, such as fal.ai, Replicate, and Black Forest Labs, using the requests library.
This behavior is integral to the skill's primary function of retrieving generated images.
The sources involved are recognized technology services.
[INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided text prompts which are subsequently processed by external image generation models.
Ingestion points: The prompt parameter in the flux_fal, flux_replicate, and flux_bfl Python functions.
Boundary markers: No specific delimiters or instructions to ignore embedded commands are present in the code snippets.
Capability inventory: The skill possesses network read capabilities (via requests.get) and filesystem write capabilities (via Path.write_bytes) to save resulting images.
Sanitization: No explicit sanitization or validation is performed on the prompt before it is sent to the providers.
Risk Assessment: The security risk is low because the prompt influences image generation results rather than the agent's execution logic, and the file write is the expected behavior for an image generation tool.

flux-model