skills/evolv3-ai/vibe-skills/admin/Gen Agent Trust Hub

admin

Pass

Audited by Gen Agent Trust Hub on Apr 10, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill makes extensive use of Bash and PowerShell scripts to perform system administration tasks.
  • Evidence includes scripts for tool installation (Initialize-AdminProfile.ps1), service management (unix.md, windows.md), and system configuration (Set-WslResources.ps1).
  • [DATA_EXFILTRATION]: Synchronizes device profiles and system metadata to a remote GitHub repository for multi-device coordination.
  • Implementation found in Sync-DeviceProfile.ps1 and sync-profile.sh using git push to a user-configured ADMIN_PROFILE_REPO.
  • Mentions integration with mem.self-host.io for semantic memory storage (references/memory-integration.md).
  • [EXTERNAL_DOWNLOADS]: Fetches sensitive configuration data and credential blobs from external secret management platforms.
  • Uses the Infisical CLI and custom scripts (Resolve-SecretRef.ps1, render-runtime.sh) to download secrets via infisical:// URIs.
  • [REMOTE_CODE_EXECUTION]: Facilitates the installation and execution of external software and MCP servers from package registries.
  • Provides instructions and scripts for installing software via npx, npm, pip, uv, winget, scoop, brew, and apt.
  • Documentation includes standard piped installation patterns for tools like Homebrew and Chocolatey (references/unix.md, references/windows.md).
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 10, 2026, 08:14 AM
Security Audit — agent-trust-hub — admin