The Agent Skills Directory

[EXTERNAL_DOWNLOADS] (HIGH): The skill clones a template from an untrusted repository: https://github.com/Eng0AI/denuvo-slides-template.git. This source is not part of the defined trusted organizations or repositories.
[COMMAND_EXECUTION] (HIGH): The skill runs pnpm install immediately after cloning the untrusted repository. This is a critical risk as pnpm will execute any pre-install, post-install, or other lifecycle scripts defined in the untrusted package.json, leading to arbitrary code execution.
[COMMAND_EXECUTION] (MEDIUM): The skill executes build and deployment commands (vercel build, pnpm build) on the unverified code, which provides further opportunities for malicious logic to execute during the build process.
[CREDENTIALS_UNSAFE] (LOW): The deployment instructions reference $VERCEL_TOKEN. While using environment variables is standard practice, the execution of untrusted code on the same system creates a risk of environment variable and token exfiltration.

denuvo-slides