The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or data exfiltration risks were identified in the skill's instructions or templates.- [COMMAND_EXECUTION]: The skill instructs the agent to use local tools rsvg-convert and asciidoctor for rendering SVGs to PNGs and compiling AsciiDoc files. These are standard documentation procedures and the use of the --safe-mode=safe flag for AsciiDoctor is a documented security best practice.- [PROMPT_INJECTION]: The skill includes security-positive instructions that explicitly prohibit embedding external assets (e.g., <image href="https://…">) in SVGs. This policy prevents potential data exfiltration or SSRF attacks that could occur if the SVG were to fetch remote resources during rendering.

ref-svg-diagrams