Merge pull request #3 from andybbruno/fix/refactor+tests

Fix/refactor+tests

Author: andybbruno

.github/workflows/ci.yml

@@ -0,0 +1,42 @@
+name: CI
+
+on:
+  pull_request:
+  push:
+    branches: [master, main]
+
+jobs:
+  lint:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: astral-sh/setup-uv@v5
+
+      - name: Install dependencies
+        run: uv sync
+
+      - name: Ruff check
+        run: uv run ruff check .
+
+      - name: Ruff format
+        run: uv run ruff format --check .
+
+  test:
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        python-version: ["3.12", "3.13"]
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: astral-sh/setup-uv@v5
+        with:
+          python-version: ${{ matrix.python-version }}
+
+      - name: Install dependencies
+        run: uv sync
+
+      - name: Run tests
+        run: uv run pytest -v

README.md

@@ -102,6 +102,43 @@ with DockerSandbox() as backend:
 print("Done!")
 ```
 
+## Example
+
+The [pizza agent](examples/pizza_agent.py) searches the web for a Neapolitan pizza recipe and writes it to a file in the workspace:
+
+```python
+from deepagents import create_deep_agent
+from deepagents_docker import DockerSandbox
+
+backend = DockerSandbox(
+    workspace_dir="examples/data",
+    allow_outbound_traffic=True,
+)
+
+agent = create_deep_agent(
+    model="openai:gpt-5.5",
+    backend=backend,
+    system_prompt="You are a pizza chef.",
+)
+
+for step in agent.stream(
+    {"messages": "Find the best neapolitan pizza recipe and write it to the recipe.md file."},
+    stream_mode="updates",
+):
+    for update in step.values():
+        if update and (messages := update.get("messages")):
+            for message in messages:
+                message.pretty_print()
+```
+
+From a clone of this repo (requires an OpenAI API key):
+
+```bash
+uv run python examples/pizza_agent.py
+```
+
+The agent writes `recipe.md` under `examples/data/`.
+
 ## Development
 
 ```bash

examples/pizza_agent.py

@@ -0,0 +1,24 @@
+from deepagents import create_deep_agent
+
+from deepagents_docker import DockerSandbox
+
+backend = DockerSandbox(
+    workspace_dir="examples/data",
+    allow_outbound_traffic=True,
+)
+
+agent = create_deep_agent(
+    model="openai:gpt-5.5",
+    backend=backend,
+    system_prompt="You are a pizza chef.",
+)
+
+if __name__ == "__main__":
+    for step in agent.stream(
+        {"messages": "Find the best neapolitan pizza recipe and write it to the recipe.md file."},
+        stream_mode="updates",
+    ):
+        for update in step.values():
+            if update and (messages := update.get("messages")):
+                for message in messages:
+                    message.pretty_print()

pyproject.toml

@@ -19,6 +19,7 @@ classifiers = [
     "Programming Language :: Python :: 3.12",
     "Programming Language :: Python :: 3.13",
     "Topic :: Software Development :: Libraries :: Python Modules",
+    "Topic :: Scientific/Engineering :: Artificial Intelligence",
 ]
 dependencies = [
     "deepagents>=0.6.7",
@@ -28,6 +29,7 @@ dependencies = [
 dev = [
     "build>=1.2.0",
     "pytest>=9.0.0",
+    "ruff>=0.9.0",
     "twine>=6.0.0",
 ]
 
@@ -41,4 +43,18 @@ packages = ["src/deepagents_docker"]
 [tool.pytest.ini_options]
 testpaths = ["tests"]
 
+[tool.ruff]
+target-version = "py312"
+line-length = 100
 
+[tool.ruff.lint]
+select = [
+    # isort
+    "I",
+
+    # Pyflakes
+    "F",
+
+    # Pyupgrade
+    "UP",
+]

src/deepagents_docker/__init__.py

@@ -1,7 +1,6 @@
 """Docker-backed sandbox backend for DeepAgents."""
 
-from deepagents_docker.backend import (
-    DockerSandbox,
-)
+from .backend import DockerSandbox
+from .errors import DockerError
 
-__all__ = ["DockerSandbox"]
+__all__ = ["DockerError", "DockerSandbox"]

src/deepagents_docker/_docker.py

@@ -7,9 +7,7 @@
 from collections.abc import Sequence
 from dataclasses import dataclass
 
-
-class DockerError(RuntimeError):
-    """Raised when a Docker CLI invocation fails."""
+from .errors import DockerError
 
 
 @dataclass(frozen=True)

src/deepagents_docker/backend.py

@@ -4,40 +4,28 @@
 
 import atexit
 import shlex
-import subprocess
 import tempfile
 import uuid
 from pathlib import Path
 
 from deepagents.backends.filesystem import FilesystemBackend
 from deepagents.backends.protocol import ExecuteResponse, SandboxBackendProtocol
 
-from deepagents_docker._docker import (
-    DockerError,
+from ._docker import (
     docker_available,
     format_docker_error,
     inspect_container_id,
     run_docker,
 )
+from .errors import DockerError
 
 DEFAULT_EXECUTE_TIMEOUT = 120
 DEFAULT_IMAGE = "python:3.12-bookworm"
 CONTAINER_WORKDIR = "/workspace"
 
 
 class DockerSandbox(FilesystemBackend, SandboxBackendProtocol):
-    """Filesystem backend with shell commands executed inside a Docker container.
-
-    File operations (`ls`, `read`, `write`, `edit`, `grep`, `glob`) run against a
-    dedicated workspace directory on the host via `FilesystemBackend` with
-    `virtual_mode=True`. The same directory is bind-mounted into the container at
-    `/workspace`, and the `execute` tool runs commands there with Docker resource
-    and security limits.
-
-    This is defense in depth, not a perfect isolation boundary. Do not mount
-    secrets into the workspace, keep Docker patched, and prefer microVMs for
-    hostile multi-tenant workloads.
-    """
+    """Docker-backed sandbox backend for DeepAgents."""
 
     def __init__(
         self,
@@ -47,8 +35,8 @@ def __init__(
         workspace_dir: str | Path | None = None,
         timeout: int = DEFAULT_EXECUTE_TIMEOUT,
         max_output_bytes: int = 100_000,
-        memory: str = "512m",
-        cpus: float = 1.0,
+        memory: str = "256m",
+        cpus: float = 0.5,
         pids_limit: int = 128,
         auto_remove: bool = True,
         extra_run_args: list[str] | None = None,
@@ -57,14 +45,14 @@ def __init__(
 
         Args:
             image: Docker image for command execution (default: official ``python:3.12-bookworm``).
+            allow_outbound_traffic: Allow/deny outbound network traffic (default: allow).
             workspace_dir: Host directory for agent files. A temporary directory is
                 created when omitted.
             timeout: Default command timeout in seconds.
             max_output_bytes: Maximum combined stdout/stderr captured per command.
-            memory: Docker memory limit (for example ``"512m"``).
+            memory: Docker memory limit (for example ``"256m"``).
             cpus: Docker CPU limit.
             pids_limit: Maximum number of PIDs inside the container.
-            outbound_traffic: Allow/deny outbound network traffic (default: allow).
             auto_remove: Remove the container on ``close()``.
             extra_run_args: Additional ``docker run`` flags appended before the image.
         """
@@ -146,9 +134,9 @@ def _start_container(self) -> None:
             "ALL",
             "--read-only",
             "--tmpfs",
-            "/tmp:rw,noexec,nosuid,size=64m",
+            "/tmp:rw,noexec,nosuid,size=512m",
             "--tmpfs",
-            "/var/tmp:rw,noexec,nosuid,size=64m",
+            "/var/tmp:rw,noexec,nosuid,size=512m",
             "-v",
             f"{self._workspace}:{CONTAINER_WORKDIR}:rw",
             "-w",
@@ -198,41 +186,41 @@ def execute(
         wrapped = self._wrap_command(command)
         docker_args = [
             "exec",
+            "-w",
+            CONTAINER_WORKDIR,
             self._container_name,
             "sh",
             "-c",
             wrapped,
         ]
 
         try:
-            completed = subprocess.run(  # noqa: S602
-                ["docker", *docker_args],
-                check=False,
-                capture_output=True,
-                text=True,
-                timeout=effective_timeout,
-            )
-        except subprocess.TimeoutExpired:
-            if timeout is not None:
-                msg = (
-                    f"Error: Command timed out after {effective_timeout} seconds "
-                    "(custom timeout). The command may be stuck or require more time."
+            completed = run_docker(docker_args, timeout=effective_timeout)
+        except DockerError as exc:
+            detail = str(exc)
+            if "timed out" in detail:
+                if timeout is not None:
+                    msg = (
+                        f"Error: Command timed out after {effective_timeout} seconds "
+                        "(custom timeout). The command may be stuck or require more time."
+                    )
+                else:
+                    msg = (
+                        f"Error: Command timed out after {effective_timeout} seconds. "
+                        "For long-running commands, re-run using the timeout parameter."
+                    )
+                return ExecuteResponse(output=msg, exit_code=124, truncated=False)
+            if "not found on PATH" in detail:
+                return ExecuteResponse(
+                    output=(
+                        "Error executing command (FileNotFoundError): "
+                        "docker executable not found on PATH"
+                    ),
+                    exit_code=1,
+                    truncated=False,
                 )
-            else:
-                msg = (
-                    f"Error: Command timed out after {effective_timeout} seconds. "
-                    "For long-running commands, re-run using the timeout parameter."
-                )
-            return ExecuteResponse(output=msg, exit_code=124, truncated=False)
-        except FileNotFoundError:
-            return ExecuteResponse(
-                output="Error executing command (FileNotFoundError): docker executable not found on PATH",
-                exit_code=1,
-                truncated=False,
-            )
-        except Exception as exc:  # noqa: BLE001
             return ExecuteResponse(
-                output=f"Error executing command ({type(exc).__name__}): {exc}",
+                output=f"Error executing command (DockerError): {exc}",
                 exit_code=1,
                 truncated=False,
             )

src/deepagents_docker/errors.py

@@ -0,0 +1,2 @@
+class DockerError(RuntimeError):
+    """Raised when a Docker CLI invocation fails."""