backtesting-frameworks
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data and model results which could contain malicious instructions. \n
- Ingestion points: External datasets retrieved via sandbase_run tools in SKILL.md. \n
- Boundary markers: The SandBase boundary section provides usage limits but lacks instructions to the agent to treat external content as untrusted data or ignore embedded instructions. \n
- Capability inventory: The skill allows file inspection and network access via vendor-specific tools (sandbase_run, sandbase_discover). \n
- Sanitization: The skill contains quality gates for data validation (units, types) but lacks sanitization for potential natural language injection within the data.
Audit Metadata