skills/openai/plugins/next-forge/Gen Agent Trust Hub

next-forge

Pass

Audited by Gen Agent Trust Hub on Sep 3, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [Remote Scaffolding Tool]: The skill documentation includes the command npx next-forge@latest init. This command downloads and executes code from a remote registry to scaffold the project structure.
  • [CLI Tool Execution]: The skill incorporates common development commands such as pnpm migrate, prisma studio, and stripe listen. These tools interact with the local development environment and external services to manage databases and webhooks.
  • [Indirect Prompt Injection Surface]: The skill monitors project files to provide assistance.
  • Ingestion points: Project configuration and source code in apps/ and packages/ folders.
  • Boundary markers: None explicitly specified for incoming project data.
  • Capability inventory: Execution of project commands like pnpm, prisma, stripe, and npx.
  • Sanitization: No explicit sanitization or filtering of project-sourced data is described.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 3, 2026, 05:56 AM
Security Audit — agent-trust-hub — next-forge