setup-review-sandbox
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEPRIVILEGE_ESCALATIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [PRIVILEGE_ESCALATION]: The skill identifies tasks requiring administrative privileges, such as installing kernel modules or container runtimes, and instructs the user to perform them manually in their terminal rather than attempting to execute them directly.
- [EXTERNAL_DOWNLOADS]: The documentation includes instructions for the user to download gVisor from its official source. This is a well-known and trusted repository for security-focused container runtimes, and the agent itself is not granted network access tools to perform these downloads.
- [COMMAND_EXECUTION]: The skill performs local command execution using docker and internal shell scripts to build and verify a toolchain image. These actions are transparently documented and consistent with the skill's stated purpose of environment setup within the vendor's ecosystem.
- [INDIRECT_PROMPT_INJECTION]: The skill's primary purpose is to provide a defense-in-depth mechanism by setting up an isolated environment (gVisor/Docker) specifically to safely handle untrusted input from pull requests, thereby reducing the overall attack surface.
Audit Metadata