review-work

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

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime flow calls MCP tools like get_work_unit, get_task, and list_task_comments to ingest outsider-authored work items/comments (e.g., issue/task bodies and progress notes written by other contributors) into the agent’s LLM context for the acceptance-criteria audit.