ai-trading-crew
Pass
Audited by Gen Agent Trust Hub on May 10, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The system is vulnerable to indirect prompt injection via the 'Sentiment & News' team which processes untrusted external data.
- Ingestion points: The skill ingests data from external news feeds, social sentiment, and analyst ratings through the Polygon.io API and NLP agents (SKILL.md).
- Boundary markers: There are no defined boundary markers or instructions to the agents to ignore or escape embedded commands within the ingested sentiment data.
- Capability inventory: The 'Execution & Ops' team has the authority to submit orders to the Alpaca broker based on the consensus reached by other agents.
- Sanitization: No sanitization or validation mechanisms are described to filter out malicious payloads hidden in financial news or social media text.
Audit Metadata