prompt-engineering

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

• Third-party content exposure detected (high risk: 0.80). The skill's required workflow and tool guidance (e.g., references/03-tool-use.md defines web_search/web_fetch and references/05-anti-hallucination.md and references/10-agentic-workflows.md explicitly instruct the agent to search and ingest public web pages and use those results to inform decisions), so the agent is expected to read untrusted, user-generated third-party web content that can materially influence subsequent tool use and actions.