Skills
Navigate
skills/bagelhole/devops-security-agent-skills/vercel-deployments/Gen Agent Trust Hub

vercel-deployments

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to install the Vercel CLI and telemetry packages (@vercel/speed-insights, @vercel/analytics) from the official npm registry.
  • [COMMAND_EXECUTION]: Numerous CLI commands are provided for deployment management, including vercel login, vercel link, vercel deploy, and environment variable operations like vercel env add and vercel env pull.
  • [REMOTE_CODE_EXECUTION]: The skill includes a GitHub Actions workflow that executes the Vercel CLI and official GitHub Actions (actions/checkout, actions/setup-node, actions/github-script) to automate previews.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external project files during the deployment pipeline.
  • Ingestion points: vercel.json configuration, middleware scripts, and application source code.
  • Boundary markers: None present in the instructional snippets.
  • Capability inventory: Shell command execution via the vercel CLI, network operations for deployments, and file system access for environment variable management.
  • Sanitization: Relies on standard Vercel CLI and platform validation.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 07:05 AM