The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell scripts and Python code for its primary purpose of diagram processing. The execution is confined to local files bundled within the skill and uses standard, well-known tools like mermaid-cli and google-chrome-stable for rendering. These actions are within the intended scope of the skill.
[REMOTE_CODE_EXECUTION]: The documentation provides standard instructions for installing dependencies like google-chrome-stable and @mermaid-js/mermaid-cli from official repositories. These are common administrative tasks and follow established software distribution norms. No arbitrary or hidden remote execution was detected.
[INDIRECT_PROMPT_INJECTION]: The extract_diagrams.py script parses Markdown files to extract Mermaid code blocks. While it processes external data, the extraction logic is restricted to specific code block patterns (```mermaid) and does not involve instructions that would influence the agent's behavior. The risk is minimized by the specialized nature of the parser.

mermaid-tools