vercel-agent-deploy-to-vercel

Pass

Audited by Gen Agent Trust Hub on Sep 2, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses various shell commands to interact with the environment, including git, vercel, and tar. These are standard tools for deployment workflows and are used as intended for project state gathering and file packaging.
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions to install the official Vercel CLI via npm install -g vercel. This is a standard installation procedure for a well-known service.
  • [DATA_EXFİLTRATION]: The deployment scripts (deploy-codex.sh and deploy.sh) package project files and upload them to Vercel deployment endpoints (codex-deploy-skills.vercel.sh and claude-skills-deploy.vercel.com). This behavior is the primary purpose of the skill. Security is maintained by explicitly excluding sensitive paths like .env, .env.*, and .git from the upload package using tar --exclude flags.
  • [SAFE]: The skill is authored by 'theangrypit' (mirrored from Vercel Labs) and all external endpoints and tools mentioned (Vercel CLI, Vercel deployment APIs) are official vendor resources belonging to Vercel. No malicious obfuscation, persistence mechanisms, or privilege escalations were detected.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 2, 2026, 12:52 PM
Security Audit — agent-trust-hub — vercel-agent-deploy-to-vercel