The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The installation instructions direct the user to clone a repository from an unverified source (https://github.com/mallahyari/system-design-visualizer.git) and execute installation and development commands (npm install, npm run dev). Running code from unverified third-party repositories poses a supply chain risk.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-uploaded images using AI to generate system diagrams and component details. This pipeline is vulnerable to indirect prompt injection, where an attacker could embed malicious instructions within a design image to manipulate the generated output or influence downstream analysis.
Ingestion points: SKILL.md (Image data processed via ImageUploader and passed to analyzeSystemDesign).
Boundary markers: Absent; the prompts to OpenAI do not include delimiters or instructions to ignore embedded text.
Capability inventory: The skill parses Mermaid diagrams and performs recursive AI calls on extracted node labels to provide component details.
Sanitization: No sanitization or validation is performed on the AI-generated Mermaid code or the component labels before they are processed by the application or subsequent AI calls.

system-design-visualizer-tool