ai-devkit-setup
Warn
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONPRIVILEGE_ESCALATIONEXTERNAL_DOWNLOADS
Full Analysis
- [PRIVILEGE_ESCALATION]: The skill instructions explicitly request the agent to execute commands outside the filesystem sandbox to modify host-level configuration, specifically targeting paths like
~/.pi/agent. This is a deliberate instruction to bypass environment security boundaries to ensure setup writes are successful. - [COMMAND_EXECUTION]: The workflow relies on executing shell commands (
ai-devkit statusandai-devkit setup) to inspect and alter the host environment. These commands process JSON output and modify local integrations. - [EXTERNAL_DOWNLOADS]: The skill uses
npx ai-devkit@latestas a fallback, which downloads and executes the latest version of the package from the public NPM registry at runtime.
Audit Metadata