wendy-contributing
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTION
Full Analysis
- EXTERNAL_DOWNLOADS (LOW): The skill provides instructions to download a container registry asset from
github.com/wendylabsinc. While the organization is not in the predefined trusted list, GitHub is a standard platform for hosting OS components, and the usage is consistent with the skill's purpose. - COMMAND_EXECUTION (LOW): Includes high-privilege commands like
sudo ddandsudo bmaptoolfor flashing disk images to hardware. These are standard procedures for embedded Linux development and are appropriately documented for their context. - REMOTE_CODE_EXECUTION (LOW): Guidance is provided to download a tarball from GitHub and import it as a container image via
ctr images import. This constitutes intended functionality for setting up a local development environment. - [Indirect Prompt Injection] (LOW): The skill describes processing device metadata (UUID and name) for mDNS discovery. There is a surface for indirect injection if these values are attacker-controlled, but the risk is low given the intended local network scope and typical OS deployment trust models.
Audit Metadata