The Agent Skills Directory

[SAFE]: The skill loads the Mermaid.js library from jsDelivr (https://cdn.jsdelivr.net/npm/mermaid@11/dist/mermaid.min.js), which is a well-known and trusted content delivery network.\n- [SAFE]: The skill uses the 'open' command via Bash to facilitate the viewing of the generated HTML file. This action is restricted to local file paths generated by the skill itself and serves the primary user-facing purpose.\n- [SAFE]: The skill exhibits a surface for indirect prompt injection because it ingests and processes content from local codebases to create diagrams. However, this is a functional requirement and no malicious bypass or exfiltration patterns were detected. Ingestion points: Codebase exploration in Step 2. Boundary markers: None. Capability inventory: Read, Write, Bash, Task. Sanitization: The skill does not explicitly describe sanitizing file content before inclusion in the HTML/Mermaid output.

visualize