personal-opsec
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXPOSURE]: The skill identifies sensitive file paths, including SSH keys, AWS credentials, and environment files. However, it includes explicit instructions to never read the contents of these files, instead directing the agent to check metadata (permissions) and match counts (grep -c) to report on the state of the system without exposing secret values.
- [COMMAND_EXECUTION]: The skill uses a variety of system utility commands (e.g.,
fdesetup,csrutil,sw_vers,scutil) to audit the security posture of the host machine. These commands are executed in a read-only capacity to report status, which aligns with the skill's primary defensive purpose. - [SAFE]: The skill establishes clear authorization and consent boundaries, explicitly requiring the agent to confirm the user's consent and scope before proceeding with the audit.
- [SAFE]: The skill provides comprehensive coverage of modern AI-specific security risks, such as indirect prompt injection in inbox assistants and trust boundaries for MCP servers, providing valuable defensive insights without employing malicious techniques.
Audit Metadata