zmm-dependency
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill defines a logical framework for business risk assessment. No malicious code, obfuscation, or unauthorized access patterns were identified.
- [DATA_EXPOSURE]: The skill accesses local files (config.yaml) and specified memory directories. These operations are restricted to the agent's operational scope and do not target sensitive system credentials or private user files.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided business scenarios, creating an ingestion surface for untrusted data. However, the agent's capabilities are limited to analytical processing and local memory storage, with no access to external networks or shell commands, effectively neutralizing injection risks.
- [CREDENTIALS_UNSAFE]: No hardcoded secrets, API keys, or credentials were found. The skill explicitly instructs the agent to inform the user if configuration files are missing rather than using default values.
Audit Metadata