Create a Daytona sandbox and run code in it with the Python or TypeScript SDK (API key auth, first execution)

Documented steps

1. Create a Daytona account and generate an API key from the Daytona dashboard API Keys page; copy the key immediately because only masked values are shown afterwards.
2. Install the SDK: Python `pip install daytona`; TypeScript via the npm package documented on the TypeScript SDK reference page.
3. Set the key as environment variable `DAYTONA_API_KEY`. Optionally set `DAYTONA_API_URL` (default `https://app.daytona.io/api`) and `DAYTONA_TARGET` (region, e.g. `us` or `eu`).
4. Initialize the client from env vars: Python `from daytona import Daytona; daytona = Daytona()`. Or explicitly: `from daytona import Daytona, DaytonaConfig; daytona = Daytona(DaytonaConfig(api_key='...', api_url='https://app.daytona.io/api', target='us'))`. TypeScript: `new Daytona({ apiKey: '...' })`.
5. For async Python use `from daytona import AsyncDaytona` with `async with AsyncDaytona(config) as daytona:`.
6. Create the sandbox with `daytona.create()` (sync) or `await daytona.create()` (async). Sandboxes default to 1 vCPU, 1GB RAM and 3GiB disk.
7. Run a code snippet statelessly: Python `response = sandbox.process.code_run('print("Hello, World!")')`; TypeScript `await sandbox.process.codeRun(code, params?, timeout?)`. Pass `params=CodeRunParams(argv=['Daytona'], env={'FOO':'BAR'})` to supply argv and environment variables.
8. Run a shell command instead: Python `sandbox.process.exec('ls -la', cwd='workspace/src', env={...}, timeout=30)`; TypeScript `await sandbox.process.executeCommand(command, cwd?, env?, timeout?)`.
9. Read `response.result` for output and check `response.exit_code` explicitly — a non-zero exit code does not always raise.
10. Clean up with `sandbox.delete()` when the work is done. Official docs: https://www.daytona.io/docs/en, https://www.daytona.io/docs/en/sandboxes, https://www.daytona.io/docs/en/process-code-execution, https://www.daytona.io/docs/en/python-sdk