alpaca-trading-backtest

Warn

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill facilitates the installation of the Alpaca CLI using 'go install' from the vendor's GitHub repository or via Homebrew. This is the primary tool used for market data acquisition.
  • [COMMAND_EXECUTION]: The skill requires a high level of environment permissions (documented as 'all' or equivalent) to allow the Alpaca CLI to access local authentication profiles and write artifacts to the filesystem.
  • [REMOTE_CODE_EXECUTION]: A central component of the skill's workflow is the agent's ability to translate user-defined trading strategies into a local Python script ('run.py') which is subsequently executed to perform the backtest simulation.
  • [COMMAND_EXECUTION]: The agent executes various shell commands including 'alpaca doctor' for connectivity checks and 'alpaca data' for fetching historical price bars, quotes, and corporate actions.
  • [DATA_EXFILTRATION]: The skill manages sensitive API credentials (ALPACA_API_KEY and ALPACA_SECRET_KEY) used for authentication with the Alpaca platform. The instructions include explicit safeguards to prevent the agent from logging or exposing these secrets in reports or shell history.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 17, 2026, 03:19 PM
Security Audit — agent-trust-hub — alpaca-trading-backtest