risk-engine
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or security vulnerabilities were detected across the skill's instructions and reference files. The provided MQL5 code snippets for risk management follow standard industry practices for algorithmic trading.
- [CREDENTIALS_SAFE]: The skill includes a dedicated safety guide (references/live-execution-safety.md) that explicitly advises against hardcoding or harvesting credentials, recommending that tokens and passwords only be obtained from environment variables or direct user input.
- [COMMAND_EXECUTION]: The skill does not contain any shell commands or system-level execution patterns. The MQL5 code is intended for use within a sandboxed trading environment.
- [DATA_EXFILTRATION]: No unauthorized network calls or file access patterns were identified. The logic focuses on internal account metrics (balance, equity) provided by the trading platform API.
- [DYNAMIC_CONTEXT_INJECTION]: The skill does not use any dynamic context injection patterns (e.g., exclamation-backtick syntax) to execute shell commands at load time.
Audit Metadata