mermaid-visualizer

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The Mermaid Visualizer runtime workflow ingests the user-provided “text content” in steps 1–5 (Analyze content → choose diagram type/configuration → generate Mermaid code → output in markdown) and uses that free text to produce diagram nodes, making it exposed to outsider-authored prompt-injection strings contained in the user’s input.