Quickly sandbox your AI agent.
Python
"""Tests for readonly mounts."""
from __future__ import annotations
import pytest
@pytest.mark.integration
@pytest.mark.slow
class TestReadonlyMount:
"""Readonly mount tests using shared sandbox."""
@pytest.mark.asyncio
async def test_readonly_can_read(self, shared_sandbox):
"""Test reading from readonly mount."""
result = await shared_sandbox.execute("cat /mnt/host/test.txt")
assert result.exit_code == 0
assert "host content" in result.stdout
@pytest.mark.asyncio
async def test_readonly_cannot_write(self, shared_sandbox):
"""Test writing to readonly mount fails (enforced by 9p and SMB)."""
result = await shared_sandbox.execute("touch /mnt/host/newfile.txt")
assert result.exit_code != 0