The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a bundled Python script to render diagrams. This script is implemented securely, using subprocess calls without a shell and validating that the rendering executable is a legitimate file.- [INDIRECT_PROMPT_INJECTION]: The agent is tasked with reading user-provided input files to ground the diagram generation. While this creates a surface for processing potentially malicious data, the skill includes explicit guardrails to prevent the agent from including executable code, scripts, or remote links in the resulting Mermaid source.- [EXTERNAL_DOWNLOADS]: The skill may fetch the Mermaid CLI from the npm registry using npx. This targets a well-known and trusted service and is a standard dependency for the skill's primary rendering functionality.

to-diagram