om-setup-agent-pipeline
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill downloads the
agent-browserbinary from Vercel Labs' official GitHub repository. This is a trusted organization, and the skill implements high-security practices by verifying the downloaded binary against a hardcoded SHA-256 checksum before execution. - [INDIRECT_PROMPT_INJECTION]: The skill analyzes repository metadata, file structures, and issue/PR descriptions to generate documentation (e.g.,
AGENTS.md,SDLC.md). While this represents an attack surface for indirect prompt injection, the skill includes explicit "Untrusted content boundary" rules and validation logic (regex-based sanitization) to ensure that external data is treated as information rather than instructions. - [COMMAND_EXECUTION]: The skill executes standard development toolchains (such as
pnpm,npm,git) and authorized tracker CLIs (gh,glab,acli,linear) to inspect the environment and manage labels. These actions are limited to the documented setup steps and are consistent with the skill's primary purpose. - [SAFE]: No malicious patterns, obfuscation, or persistence mechanisms were detected. The skill follows best practices for secret hygiene and autonomous execution boundaries.
Audit Metadata