CodeRunner FAQs

Question 1

Can I process my local files (images, PDFs, videos) with AI using CodeRunner?

Accepted Answer

Yes, a core use case for CodeRunner is enabling LLMs to directly process your local files (e.g., videos, images, documents, data) by generating and executing code within the secure sandbox on your Mac. This allows for analysis, transformation, or processing without files leaving your machine.

Question 2

Which LLM platforms and integrations does CodeRunner support?

Accepted Answer

CodeRunner offers broad integration with popular LLM platforms including Claude Desktop, OpenAI Agents (Python library), Google's Gemini-CLI, Amazon's Kiro, and its own Coderunner-UI. It acts as a universal MCP server for various AI agents.

Question 3

What are the system requirements to run CodeRunner?

Accepted Answer

CodeRunner requires a Mac running macOS with an Apple Silicon processor (M1/M2/M3/M4) and Python 3.10 or newer installed. It's designed to take full advantage of Apple's native container technology.

Question 4

How does CodeRunner ensure data privacy and security?

Accepted Answer

CodeRunner utilizes Apple containers for VM-level isolation, ensuring code runs in a highly secure sandbox separate from your host system. Its key benefit is processing your local files with remote LLMs without requiring cloud upload, keeping your sensitive data private on your machine.

Question 5

What is CodeRunner and how does it work?

Accepted Answer

CodeRunner is an MCP (Model Context Protocol) server that executes AI-generated code in a secure, local sandbox environment on your Mac, leveraging Apple's native containers for robust isolation. This allows LLMs to process your local data directly on your machine.

CodeRunner

CodeRunner

Key Features

Use Cases

Key Features

Use Cases