hzdb-cli
Pass
Audited by Gen Agent Trust Hub on Mar 8, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill facilitates the installation of the
@meta-quest/hzdbpackage via npm. These resources originate from the verified vendor 'meta-quest' and are handled as trusted components. - [COMMAND_EXECUTION]: The skill utilizes the
Bash(hzdb:*)tool to execute bridge commands. This includes powerful capabilities such as installing/uninstalling applications, managing files on the device filesystem, and executing arbitrary commands via the device shell. - [PROMPT_INJECTION]: The skill contains an indirect prompt injection surface. It ingests untrusted data from external sources including device logs (via
hzdb logandhzdb adb logcat) and developer documentation (viahzdb docs fetch). There are no explicit boundary markers or 'ignore' instructions provided to the agent to mitigate potential instructions embedded in these external data streams. The capability to execute shell commands and install apps increases the potential impact if the agent were to follow instructions found within the ingested data.
Audit Metadata