rhino-sdk-harmonize
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill instructions and code snippets describe legitimate usage of the Rhino Health SDK for data processing and harmonization. The content is educational and aligns with the stated purpose of the skill.
- [DYNAMIC_EXECUTION]: The documentation identifies ROW_PYTHON and TABLE_PYTHON transformation types. These are standard features of the SDK that allow for custom logic during the data transformation process.
- [INDIRECT_PROMPT_INJECTION]: The skill describes data harmonization pipelines which ingest external datasets and use LLM-based configuration generation, creating a potential surface for indirect prompt injection from untrusted data.
- Ingestion points: input_dataset_uids in DataHarmonizationRunInput (SKILL.md).
- Boundary markers: None mentioned in the instructions.
- Capability inventory: SDK methods for creating mappings and running harmonization, including session.syntactic_mapping.run_data_harmonization (SKILL.md).
- Sanitization: No sanitization or validation of source data is explicitly mentioned in the guide.
Audit Metadata