humaans-io

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

• Third-party content exposure detected (medium risk: 0.65). At runtime the skill calls Apideck’s HRIS unified API (and possibly the Proxy API) to fetch Humaans tenant data; that downstream response is free-form text originating from the outsider Humaans service and can be ingested into the agent’s LLM context.