The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a Python script (scripts/render_diagram.py) that uses subprocess.run to execute external diagram rendering engines such as mmdc (Mermaid), dot (Graphviz), and plantuml. This is the intended core functionality of the skill for generating image artifacts.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted user data (notes, code, schemas) to generate diagrams, which represents a potential attack surface.
Ingestion points: User-provided natural language, notes, code, and schemas (documented in SKILL.md).
Boundary markers: No specific delimiters or safety instructions are used to wrap external content.
Capability inventory: The skill can write files and execute subprocesses via the rendering script.
Sanitization: No explicit sanitization of the input data is performed before it is passed to rendering tools.
[EXTERNAL_DOWNLOADS]: The documentation suggests installing the @mermaid-js/mermaid-cli package via NPM and provides links to official Graphviz and PlantUML download pages. These are well-known, trusted software sources used for the skill's primary purpose.

diagram-generator