ai-artist

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly describes using web-search tools and search grounding (e.g., references/advanced-techniques.md "Agent Prompting" lists a search(query): Search the web tool, references/nano-banana.md mentions Search Grounding with google_search, and references/llm-prompting.md shows RAG retrieved_context), so the agent is expected to fetch and consume open/public third-party content that could carry indirect prompt injections.