stock-technical

Warn

Audited by Gen Agent Trust Hub on Sep 20, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the user to install third-party Python libraries including czsc and smartmoneyconcepts. These are specialized libraries from external sources that are not part of the standard Python environment or a verified organization list.
  • [COMMAND_EXECUTION]: The skill heavily utilizes shell command execution via the longbridge CLI to fetch market data, account balances, and order history. It pipes this output directly into Python execution environments.
  • [DATA_EXFILTRATION]: The skill accesses highly sensitive user information including total account net assets (longbridge assets), current stock holdings (longbridge positions), and private trading history (longbridge order). While the skill processes this data for the user, accessing such sensitive financial data represents a significant security surface.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external, untrusted data sources including stock-specific news (longbridge news), company filings (longbridge filing), and general web search results. These sources could be used to embed malicious instructions that might influence the agent's analysis or recommendations.
  • [DYNAMIC_EXECUTION]: The skill involves executing dynamic Python scripts (specifically scripts/signal_engine.py and various python3 -c one-liners) to process and analyze the stock market data fetched via the CLI.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 20, 2026, 01:13 AM