vercel-open-agents
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references downloads from Vercel Labs' official GitHub repository (
github.com/vercel-labs/open-agents) and recommended development tools likebun. These are well-known, trusted sources for the intended purpose of the skill. - [COMMAND_EXECUTION]: The skill provides standard setup commands (e.g.,
bun install,openssl rand,git commit) for local development and environment configuration. These are expected for a developer-centric skill and do not perform suspicious background operations. - [DATA_EXPOSURE]: The documentation correctly identifies sensitive environment variables (e.g.,
GITHUB_APP_PRIVATE_KEY,POSTGRES_URL) and explicitly instructs users to store them securely in.envfiles or Vercel's environment settings rather than hardcoding them, following security best practices. - [DYNAMIC_EXECUTION]: The skill describes a sandbox architecture where an agent can execute shell commands (
sandbox.exec). While this allows code execution, the skill's design emphasizes isolation via Vercel sandboxes to protect the host environment.
Audit Metadata