mmt-trading-systems
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [Prompt Injection] (SAFE): No attempts to override agent behavior or bypass safety filters were detected. The content is purely instructional and technical.
- [Data Exposure & Exfiltration] (SAFE): API keys are handled via placeholders. All network communication is directed to the legitimate MMT service endpoint (mmt.gg). File system access is restricted to saving and loading local bot state, which is standard for the described use case.
- [Unverifiable Dependencies] (SAFE): The skill utilizes standard, reputable libraries for networking and asynchronous processing. No remote script execution or unverified package installations are present.
- [Indirect Prompt Injection] (SAFE): Although the skill processes external market data, it does so within a structured programmatic context and does not exhibit vulnerability to instruction injection via data feeds.
- [Obfuscation] (SAFE): All files consist of plain, human-readable markdown and source code with no evidence of encoding or hidden content.
Audit Metadata