tracking-marketing-metrics

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

• Third-party content exposure detected (high risk: 0.80). This skill's workflow (workflow.md, Phase 2 "Self-Reported Attribution" and "AI-Specific Attribution") explicitly requires ingesting unstructured, user-generated open-text attribution answers and auditing AI/model outputs (e.g., "monthly audit of top 50 prompts across ChatGPT, Claude, Perplexity, Gemini"), which are third‑party/untrusted inputs the agent must read and that directly feed attribution decisions and budget/resource actions.