python-on-vercel

Pass

Audited by Gen Agent Trust Hub on Aug 20, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • Remote Instruction Ingestion: The skill fetches course materials and instructions from https://vercel.com/academy/. While this involves processing external content, the sources are official Vercel domains associated with the skill's author.
  • Command Execution for Progress Verification: The skill utilizes shell commands like vercel whoami, vercel --version, and fastapi dev to verify the student's environment and project state. These are standard tools for the Python on Vercel workflow.
  • Project File Inspection: To provide tailored guidance, the skill reads local configuration files such as pyproject.toml, package.json, and .vercel/project.json. This access is scoped to the student's project directory for educational purposes.
  • Instruction Following from External Content: In Teaching Mode, the agent is instructed to follow <agent-instructions> blocks contained within fetched Markdown files. This mechanism allows the skill to provide dynamic, step-by-step lessons based on live content from the Academy API.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 20, 2026, 08:37 PM
Security Audit — agent-trust-hub — python-on-vercel