The Agent Skills Directory

[SAFE]: The skill interacts with the Luma AI API (api.lumalabs.ai), which is a well-known service. All network operations are intended for image generation requests and results retrieval.
[COMMAND_EXECUTION]: The skill executes a bundled Python script (scripts/luma_imagegen.py). The script is written using standard Python libraries and does not perform any unauthorized command execution, system modifications, or privilege escalation.
[EXTERNAL_DOWNLOADS]: The script downloads generated image files from the URLs returned by the Luma API to a local output/luma/ directory. This behavior is consistent with the skill's primary purpose and targets a recognized service.
[PROMPT_INJECTION]: The skill utilizes prompt augmentation techniques to structure user requests for better image quality. There is no evidence of instructions designed to bypass safety filters, extract system prompts, or override agent behavior.
[SAFE]: API key management is handled securely by checking for the LUMA_API_KEY in the execution environment or a local .env file, avoiding the exposure of credentials in the interaction history.

luma-imagegen