The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references and downloads the official @cloudflare/sandbox package from the NPM registry. This is a standard and safe dependency for the intended purpose of building sandboxed applications on Cloudflare Workers.
[COMMAND_EXECUTION]: The instructions include standard development commands like npm install and docker info. These are required for the local development environment and do not pose a security risk in this context.
[REMOTE_CODE_EXECUTION]: The skill describes an SDK whose primary purpose is to provide an environment for executing code (sandbox.exec, sandbox.runCode). The provided examples and documentation correctly illustrate how to use these tools for building secure, isolated interpreters for AI-generated or untrusted code. All implementation patterns follow best practices for sandboxing.
[DATA_EXFILTRATION]: There are no patterns suggesting unauthorized data access or exfiltration. The skill focuses on legitimate file and network operations within a controlled sandbox environment.
[PROMPT_INJECTION]: The skill contains technical instructions for configuring AI agents to use the Sandbox SDK. It does not include instructions that attempt to bypass AI safety filters or override system prompts.

sandbox-sdk