sap-cloud-sdk-ai-python
Pass
Audited by Gen Agent Trust Hub on Aug 7, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: Indirect Prompt Injection Surface
- Ingestion points: The skill processes untrusted user data via placeholder values in orchestration templates (e.g.,
{{?question}},{{?text}}) and chat completion messages. - Boundary markers: The skill uses double-brace templating syntax (
{{?placeholder}}) to delimit dynamic content. - Capability inventory: Includes capabilities for network API calls to SAP AI Core and model providers, file system interaction via configuration files, and tabular data prediction.
- Sanitization: The skill explicitly documents and provides examples for using SAP BTP's Orchestration Service features for safety, including Azure Content Safety filters and Data Privacy Integration (DPI) for data masking.
- [EXTERNAL_DOWNLOADS]: The skill references an external repository for the Databricks agent skills plugin on GitHub (
github.com/databricks/databricks-agent-skills). This is documented as an optional extension for specific environments and includes instructions to never install it without user consent.
Audit Metadata