The Agent Skills Directory

[SAFE]: No security issues detected. The skill follows standard development and security best practices.- [COMMAND_EXECUTION]: The skill uses standard command-line tools such as git, pnpm, and node to manage the project, build dependencies, and execute evaluations. These are expected behaviors for a development-focused tool.- [EXTERNAL_DOWNLOADS]: Interactions with external services are limited to official GitHub repositories for code access and Vercel APIs for cloud sandbox provisioning. These are trusted services in this context.- [CREDENTIALS_UNSAFE]: The skill correctly implements secret management by advising users to store sensitive API tokens for Vercel and Axiom in a local .env.local file rather than hardcoding them.

rde-eval