agent-orchestration-planner

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly includes a "search_web" tool in TOOL_ORDER and executes tool results which are appended to memory (see self.memory append and state.get_context used in build_planning_prompt), so it ingests open web content (untrusted third-party data) into LLM prompts, enabling indirect prompt injection.