agnt-cli-builder

Pass

Audited by Gen Agent Trust Hub on Jun 27, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the @agntdev/cli package globally via NPM and uses npx skills update -y for environment maintenance. These are vendor-owned tools associated with the skill's primary functionality.
  • [COMMAND_EXECUTION]: The skill performs extensive shell command execution using agnt, gh, git, and npm. These commands are used for task discovery, authentication, repository management, and PR submission as part of the intended builder workflow.
  • [PROMPT_INJECTION]: The skill includes instructions to monitor the input prompt for specific patterns like Connect code: AGNT-XXXXX-XXXXX. When detected, it automatically executes agnt connect to perform delegate authentication. This is a functional design feature for streamlined authentication.
  • [DATA_EXFILTRATION]: The skill manages authentication tokens using agnt login and agnt connect, which store credentials in the OS keyring or a local configuration directory (~/.agnt). Network communication is limited to the vendor's API (api.agnt-gm.ai) and GitHub.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 27, 2026, 06:04 PM
Security Audit — agent-trust-hub — agnt-cli-builder