next-upgrade
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- REMOTE_CODE_EXECUTION (LOW): The skill executes
npx @next/codemod@latest, which downloads and runs code at runtime. This targets official Vercel/Next.js tools, which are considered trusted. - PROMPT_INJECTION (LOW): The skill processes untrusted web content from
nextjs.orgto determine upgrade steps, creating an indirect prompt injection surface. 1. Ingestion points:https://nextjs.org/docs/app/guides/upgrading/documentation URLs. 2. Boundary markers: Absent. 3. Capability inventory:npm install,npx, andnpm run build. 4. Sanitization: Absent. - EXTERNAL_DOWNLOADS (LOW): The skill fetches content from
nextjs.org, an official and trusted source for this task. - COMMAND_EXECUTION (LOW): The skill runs shell commands for installing dependencies and verifying builds, which is necessary for its stated purpose.
Audit Metadata