ghost-auto-trader

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The architecture describes a pipeline that processes external TradingView signals and news sentiment data to validate and execute live financial trades.
  • Ingestion points: Signal payloads from TradingView webhooks and external news sentiment feeds are passed to the AI Gate (SKILL.md, README.md).
  • Boundary markers: The design lacks explicit delimiters or instructions to treat external data as untrusted, increasing the risk of the LLM obeying instructions embedded in signals or news.
  • Capability inventory: The system automatically executes 0DTE options trades via broker APIs (e.g., Tradier) based on the AI Gate's approval.
  • Sanitization: There are no mentioned mechanisms for sanitizing or filtering input data before it reaches the AI validation step.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 09:55 PM
Security Audit — agent-trust-hub — ghost-auto-trader