Configuring Custom Environments

Set up Python, Node.js, system packages, environment variables, and encrypted secrets. Make your agent's cloud computer work exactly the way you need it.

Your Agent's Cloud Computer

Every environment in Computer Agents is an isolated cloud computer. You control the runtime, the packages, the secrets, and the tools. This tutorial shows you how to configure it all.

Runtimes

Runtimes define the base software stack. You can enable multiple runtimes on a single environment.

Available runtimes:

Python 3.12
Node.js 20
Go 1.22

To set runtimes, go to your environment settings or use the API:

TypeScript

example.ts

Python

example.py

Packages

Install packages that your agent needs. These persist across sessions — install once, use forever.

TypeScript

example.ts

Python

example.py

Or add them in the UI under Environment > Packages.

Environment Variables

Set variables that your agent can access at runtime:

TypeScript

example.ts

Python

example.py

Encrypted Secrets

For sensitive values like API keys and tokens, use encrypted secrets. These are stored securely and never exposed in logs:

TypeScript

example.ts

Python

example.py

MCP Servers

Extend your agent's capabilities with Model Context Protocol servers:

TypeScript

example.ts

Python

example.py

Putting It All Together

Here's a complete environment configuration for a data science agent:

TypeScript

example.ts

Python

example.py

Your environment is now fully configured. Every thread created in this environment will have access to Python, all installed packages, and your database connection.

Ready to get started?

Try Computer Agents today and experience the future of AI-powered automation.

Get Started