approve-exempt

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

• Third-party content exposure detected (high risk: 0.75). The runtime workflow calls harness_list to fetch the outsider-authored pending exemption queue (e.g., reason, issue_title, requested_by, etc. originate from other users’ exemption requests) and then renders those fields into the agent’s LLM context for the chat table.