nemoclaw-user-configure-security
Pass
Audited by Gen Agent Trust Hub on Jun 18, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill consists exclusively of reference documents, metadata, and evaluation reports. It contains no executable scripts, shell commands, or active logic that could be abused.
- [PROMPT_INJECTION]: Documentation in
references/openclaw-controls.mdmentions prompt injection patterns such as "ignore all previous instructions" within a table describing system detection capabilities. This is purely educational documentation and does not constitute a functional injection attack. - [DATA_EXFILTRATION]: Technical documentation in
references/best-practices.mdanalyzes data exfiltration risks in the context of network policy configuration. These descriptions are part of a risk framework intended to help users justify and maintain a secure posture. - [CREDENTIALS_UNSAFE]: The documentation in
references/credential-storage.mddescribes secure credential management practices, such as storing secrets in the OpenShell gateway instead of on disk. No hardcoded credentials or unsafe secret handling instructions were identified.
Audit Metadata