agentic-development

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

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime workflow is not fully specified, but it explicitly includes web research patterns (e.g., web_search / read_webpage tools that fetch and extract arbitrary webpage text), which would feed outsider-authored free text from public web pages into the agent’s LLM context via tool results.