generate-rap-service-researched

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

• Third-party content exposure detected (medium risk: 0.30). The skill’s runtime workflow performs outsider-influenced LLM reads from the target SAP system by searching and reading existing user-authored RAP artifacts (e.g., SAPSearch(query="BDEF Z*"... ) / SAPRead(type="BDEF|DDLS|SRVB|...") and also reads DDL/diagnostics error text), meaning free text authored elsewhere in the same tenant can be ingested.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 1.00). The skill instructs the agent to create, modify, activate, publish, and manage transport-backed artifacts on the target SAP system (SAPWrite / SAPActivate / SAPTransport calls), which directly changes system state and can have privileged effects.