concept-builder

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

• Third-party content exposure detected (high risk: 0.85). The required runtime workflow includes mandatory web search in Step 2, which can fetch public web pages and ingest their readable text into the agent’s LLM context (public web content fetched at runtime → web pages/scraped snippets → LLM context), creating an indirect prompt-injection risk.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The repo's mermaid validation steps call npx @mermaid-js/mermaid-cli (i.e. the mermaid CLI fetched from npm: https://www.npmjs.com/package/@mermaid-js/mermaid-cli) at runtime to render/validate mermaid blocks, which downloads and executes remote code that the workflow relies on.