build-trading-strategies
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill utilizes tools (
create_strategy,create_prediction_market_strategy) that accept natural language descriptions to generate executable code, creating a surface for indirect prompt injection. \n - Ingestion points: The
descriptionparameter of the generation tools specified in SKILL.md. \n - Boundary markers: The skill includes a manual 'Validation Checklist' and warnings to test code before deployment, though no programmatic boundary enforcement is present. \n
- Capability inventory: The generated code is designed to run within the Jesse trading framework and interacts with other high-capability tools mentioned like
deploy-live-trading. \n - Sanitization: There is no evidence of input validation or sanitization for the user-provided descriptions before they are processed by the AI generation tool.
Audit Metadata