The Agent Skills Directory

[Command Execution] (LOW): The skill requests Bash(npx:*) and Bash(mmdc:*) permissions in its allowed-tools configuration. This is used to run the Mermaid CLI and Playwright-based rendering tools. While these are appropriate for the skill's stated purpose, the npx:* wildcard allows for the execution of any arbitrary Node.js package.
[External Downloads] (LOW): To function as described, the skill downloads and executes packages from the NPM registry at runtime using npx. This creates a dependency on external infrastructure (NPM) and the integrity of the downloaded packages.
[Indirect Prompt Injection] (LOW): The skill's primary function is to process and render Mermaid diagram syntax, which may include untrusted input from users or external files.
Ingestion points: Mermaid diagram definitions passed to the rendering tools (e.g., via mmdc).
Boundary markers: Absent; the skill processes raw syntax strings without explicit sanitization or delimiters described in the provided files.
Capability inventory: Includes Bash command execution and network access (via Playwright/npx) to render diagrams.
Sanitization: Not explicitly implemented in the skill; relies on the default security levels of the Mermaid rendering engine.

moai-library-mermaid