copywriting

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

• Third-party content exposure detected (high risk: 0.85). SKILL.md requires pulling personalization material via ColdIQ MCP calls (e.g., mcp__coldiq__extract_post_engagement and mcp__coldiq__fetch_page_content) which fetches/parses outsider-authored public web text at runtime and feeds it into the agent context.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly calls ColdIQ MCP inputs (mcp__coldiq__*) at runtime and directs running "COLDIQ_API_KEY= npx -y @coldiq/mcp@latest", and it uses the ColdIQ API endpoint https://api.coldiq.com to fetch personalization data that will be injected into prompts, so an external endpoint is used at runtime to supply content that directly controls the agent's output.