The Agent Skills Directory

[SAFE]: The skill performs its functions using trusted third-party libraries including Httpx for networking, and Pillow and NumPy for image manipulation. These are industry-standard packages for these tasks.- [COMMAND_EXECUTION]: The script scripts/research_image.py automates the generation of infographics by calling scripts/text_to_image.py. This is done safely via subprocess.run with an argument list, which prevents shell-related injection attacks.- [EXTERNAL_DOWNLOADS]: The skill is designed to interact with external APIs at api.openai.com and llm.api.zyuncs.com to process image data. These network operations are intrinsic to the skill's purpose and do not involve downloading or executing untrusted code.- [DATA_EXFILTRATION]: Source code review shows that network activity is confined to standard API requests for image processing. There are no attempts to access or transmit sensitive system files, environment secrets, or user credentials.

image-service