The Agent Skills Directory

[SAFE]: The skill's functionality is consistent with its stated purpose of brand research and asset management. No malicious behaviors such as obfuscation, privilege escalation, or persistence were detected.\n- [COMMAND_EXECUTION]: The skill utilizes several internal Python scripts (scaffold_brand.py, fetch_asset.py, render_product_shot.py, verify_pack.py) to automate project setup and asset management. These scripts are run locally as intended parts of the workflow.\n- [EXTERNAL_DOWNLOADS]: The skill fetches brand assets from external URLs discovered during research and downloads generated images from the FAL.ai platform using the standard Python library.\n- [DATA_EXFILTRATION]: The skill includes functionality to upload local reference images to the FAL.ai service via the fal-client library. This data transfer is a documented and necessary component of the skill's image generation feature and targets a well-known service.\n- [CREDENTIALS_UNSAFE]: The skill manages external API keys (FAL, ElevenLabs) by instructing users to store them in a .env file, which is then loaded by the scripts. This is a standard and recommended practice for secret management.

brand-research