fix(code_executors): harden ContainerCodeExecutor sandbox by default

[adilburaksen]

Summary

ContainerCodeExecutor runs model-generated code, which can be influenced by untrusted input (e.g. via prompt injection). It starts the container with default Docker networking and no capability restrictions, so the executed code can reach the cloud metadata endpoint (169.254.169.254) — which yields the host service-account token — reach internal services, or escalate privileges.

This is inconsistent with the isolation posture of every other ADK code executor:

• GkeCodeExecutor runs under gVisor with cap_drop: ["ALL"], non-root, read-only root filesystem, and a strict security context.
• BuiltInCodeExecutor / VertexAiCodeExecutor / AgentEngineSandboxCodeExecutor run in managed server-side sandboxes.
• UnsafeLocalCodeExecutor is explicitly documented as unsafe.

ContainerCodeExecutor was the only executor running code with full network access and no isolation flags or warning.

Change

• Start the container with network_disabled=True by default. This is exposed as a configurable network_disabled field — set it to False to re-enable networking when the executed code is trusted.
• Drop all Linux capabilities (cap_drop=["ALL"]) and forbid privilege escalation (security_opt=["no-new-privileges"]), matching GkeCodeExecutor.
• Document the security posture in the class docstring and point users to the sandboxed executors for untrusted code.
• Add unit tests covering the hardened defaults and the opt-in network path.

Compatibility

Code that legitimately needs network access can opt back in with ContainerCodeExecutor(..., network_disabled=False). Dropping capabilities and no-new-privileges do not affect normal Python code execution.

[adk-bot]

Response from ADK Triaging Agent

Hello @adilburaksen, thank you for creating this PR to harden the ContainerCodeExecutor sandbox!

While checking this PR against our contribution guidelines, I noticed a few things that need to be addressed:

1. Testing Plan: Please include a dedicated testing plan section in your PR description explaining how you tested these changes.
2. Logs / Test Results: Please provide logs or test output (e.g., the output of the new pytest unit tests) showing that the test suite passes successfully.
3. CI Header Check Failure: The header-check CI status check failed. Please ensure all new and modified files have the correct Apache 2.0 license headers. You can automatically verify and fix this by running pre-commit run --all-files locally before pushing.

These steps will help the maintainers review your contribution more quickly and efficiently. Thank you!

[rohityan]

Hi @adilburaksen, Thank you for your contribution! We appreciate you taking the time to submit this pull request. Can you please fix the failing unit tests before we can proceed with the review.