designmdme-cli

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

• Third-party content exposure detected (high risk: 0.85). The required workflow involves fetching a user-supplied public website URL at runtime and scanning it to generate a DESIGN.md, so the LLM context can include outsider-authored free text extracted from that external site (public web content → streamed spec/Markdown into the agent).

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

• Potentially malicious external URL detected (high risk: 0.90). The CLI fetches the target website at runtime (e.g., "stripe.com" / "https://stripe.com"), converts that remote content into a DESIGN.md which the agent is explicitly instructed to read and use to drive code generation, so external site content directly controls agent prompts/behavior.