admin
Warn
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
- [COMMAND_EXECUTION] (MEDIUM): The skill sources and executes local shell and PowerShell scripts (e.g., 'load-profile.sh', 'Initialize-AdminProfile.ps1') to configure environment variables and system state.
- [DATA_EXFILTRATION] (MEDIUM): The documentation explicitly lists 'finding configs, keys' and loading 'secrets' from '.env.local' references as core functionalities. This constitutes high-risk sensitive file access that exposes credentials to the agent's context.
- [INDIRECT_PROMPT_INJECTION] (LOW): The skill processes untrusted external strings (README instructions) to generate adapted commands. Ingestion points: System READMEs and local profiles. Boundary markers: None specified in documentation. Capability inventory: Full command execution, script sourcing, and SSH access. Sanitization: Not documented; the skill relies on string transformation logic to adapt external commands.
Audit Metadata