The Agent Skills Directory

[SAFE]: The skill instructions and associated scripts do not exhibit any malicious behaviors. All operations are confined to the agent's environment, using local assets and internal tools for image generation.- [COMMAND_EXECUTION]: The skill uses a local script scripts/check_cover.py to validate image outputs. This script is part of the skill package and uses standard libraries (Pillow) to verify aspect ratios, dimensions, and file sizes, as well as to generate preview thumbnails.- [PROMPT_INJECTION]: The skill processes user-supplied content such as titles and images. It mitigates potential indirect prompt injection by using structured templates and instructing the agent to perform visual verification and manual corrections on the output.

space-xhs-cover