The Agent Skills Directory

[SAFE]: The skill exhibits a high level of security engineering by implementing a robust validation layer for SVG content. The validate_svg function in scripts/run_expert_mvp.py explicitly blocks high-risk features such as external URLs, scripts, CSS styles, and embedded data URIs, mitigating potential exfiltration or injection vectors.
[COMMAND_EXECUTION]: The runner scripts utilize subprocess.run to call standard system utilities like resvg for image rendering and ffprobe for media metadata extraction. These operations are performed within the scope of the skill's operational directory and do not expose the system to arbitrary command injection.
[SAFE]: All external dependencies and vendored components are established tools for image and presentation processing. The setup instructions reference well-known services like Homebrew for utility installation.

happy-figure-edit