Yash Shah a9d1c6554b feat(v0.7): Docker sandbox backend for real isolation
Context:
- Subprocess + netns probes are not enough for production agent code execution.
- Ephemeral docker run --rm with --network=none and optional --memory is the
  essential isolation path for agentbox-backed products like agentforge.

Test Plan:
- docker compose run --rm test
- ruff check src tests
- pytest tests/test_docker.py (mocks CLI; live skip without docker)

Additional Notes:
- Default remains subprocess for local/CI unit tests.
- Set AGENTBOX_SANDBOX_BACKEND=docker (plus Docker CLI/daemon) for production.
2026-09-14 23:49:33 +05:30

agentbox

Self-hosted code execution sandbox for AI agents — one docker compose up gives you an HTTP API for running untrusted code in isolated environments.

PyPI License: MIT Python 3.11+ CI

Status: v0.7 — Python + Node execution, Docker sandbox backend (recommended isolation), subprocess default for local tests, timeouts, memory limits, Linux netns egress deny, TypeScript client, workspace snapshots.

60-second try

pip install agentbox-sandbox
agentbox serve   # API on :8080
# or: docker compose up agentbox
curl -s http://localhost:8080/health
curl -s -X POST http://localhost:8080/v1/run \
  -H 'Content-Type: application/json' \
  -d '{"code":"print(sum(range(10)))"}'

Why this vs alternatives

Approach Strength Gap
agentbox Self-hosted HTTP API + SDK; Docker or subprocess backends Not gVisor/Firecracker (yet)
Hosted sandboxes (E2B, etc.) Strong isolation, managed Per-second cost; data leaves your network
Raw docker exec Familiar No agent-oriented API / snapshots / limits
YOLO in the agent process Zero infra Full host compromise risk

Problem

Every agent that writes and runs code needs a safe execution environment. Teams either YOLO in shared containers or pay per-second for hosted sandboxes. Self-hosting gVisor/Firecracker is weeks of work.

Key features (v0.7)

  • HTTP API: POST /v1/run executes Python or JavaScript
  • Docker backend (AGENTBOX_SANDBOX_BACKEND=docker): ephemeral docker run --rm, --network=none, optional --memory, workspace at /work
  • Subprocess backend remains the default for easy local tests
  • Per-request limits.timeout_seconds (HTTP 408 on timeout)
  • Per-request limits.memory_mb (Docker --memory or subprocess RLIMIT_AS); response includes limits_applied + oom_killed
  • AGENTBOX_MAX_MEMORY_MB clamps requested memory
  • Default-deny egress: Docker uses --network=none; subprocess uses Linux unshare/bwrap when available (network_isolated in response)
  • Workspace snapshots: "snapshot": true then "snapshot_id"
  • Python SDK + TypeScript client (sdk/ts/client.ts)
  • Credential stripping when the backend is not unrestricted

Architecture

Agent / SDK
    └── POST /v1/run
            ├── SubprocessSandbox (default — easy tests)
            └── DockerSandbox (recommended — stronger isolation)
Component Technology Why
API FastAPI Async-ready, OpenAPI docs, widely adopted
Server uvicorn Standard ASGI server
Config pydantic-settings Typed env config
Isolation Docker / subprocess Docker for production; subprocess for CI/dev
Tests pytest + httpx TestClient Fast API testing

Installation

pip install agentbox-sandbox
pip install -e ".[dev]"

Usage

Start server

agentbox serve
# or
docker compose up agentbox

Requires the Docker CLI (and a reachable daemon) on the host running agentbox:

export AGENTBOX_SANDBOX_BACKEND=docker
# optional:
# export AGENTBOX_DOCKER_IMAGE=python:3.12-slim
# export AGENTBOX_DOCKER_NODE_IMAGE=node:20-slim
agentbox serve

Each /v1/run starts an ephemeral container, mounts a temp workspace at /work, applies timeout on docker run, and removes the container (--rm). Health and run responses report backend: "docker".

Run code

curl -X POST http://localhost:8080/v1/run \
  -H 'Content-Type: application/json' \
  -d '{"code": "print(sum(range(10)))"}'

curl -X POST http://localhost:8080/v1/run \
  -H 'Content-Type: application/json' \
  -d '{"code": "x = bytearray(10**9)", "limits": {"memory_mb": 64, "timeout_seconds": 5}}'

Python SDK

from agentbox.sdk.client import AgentboxClient

client = AgentboxClient("http://localhost:8080")
print(client.health())
print(client.run("print('hello')"))
print(client.run("console.log('hello')", language="javascript", timeout_seconds=5))
print(client.run("x = bytearray(10**8)", memory_mb=64))
snap = client.run("open('memo.txt','w').write('kept')", snapshot=True)
print(client.run("print(open('memo.txt').read())", snapshot_id=snap["snapshot_id"]))
client.close()

Docker

docker compose up agentbox        # start API on :8080
docker compose run --rm test    # run unit tests

Note: using AGENTBOX_SANDBOX_BACKEND=docker from inside a container needs Docker socket access (DinD / mounted socket). Prefer running the API on the host with the Docker backend, or keep the compose service on subprocess.

Configuration

Variable Default Description
AGENTBOX_HOST 0.0.0.0 Bind host
AGENTBOX_PORT 8080 Bind port
AGENTBOX_DEFAULT_TIMEOUT_SECONDS 30 Execution timeout
AGENTBOX_DEFAULT_MEMORY_MB unset Optional default memory cap
AGENTBOX_MAX_MEMORY_MB 8192 Clamp requested memory_mb to this max
AGENTBOX_SANDBOX_BACKEND subprocess subprocess | docker | unrestricted
AGENTBOX_DOCKER_IMAGE python:3.12-slim Image for Python runs (docker backend)
AGENTBOX_DOCKER_NODE_IMAGE node:20-slim Image for JavaScript runs (docker backend)
AGENTBOX_SNAPSHOT_DIR /tmp/agentbox-snapshots Workspace snapshot store

Running tests

pytest tests/ -v
# Docker unit tests mock the CLI; live Docker run is skipped if docker is unavailable
pytest tests/test_docker.py -v

Roadmap

  • Node.js runtime + TypeScript client + per-run timeout
  • Filesystem snapshot/restore (tar workspaces)
  • limits.memory_mb via RLIMIT_AS / Docker --memory
  • Default-deny egress via Linux netns (unshare/bwrap) when available
  • Docker ephemeral-container backend
  • gVisor runsc backend with warm pool
  • Egress allowlists (beyond all-or-nothing netns)

License

MIT

Known limitations (v0.7)

  • Subprocess default is convenient for tests — not production-grade isolation; use AGENTBOX_SANDBOX_BACKEND=docker for stronger isolation
  • Docker backend needs a local Docker CLI/daemon; missing CLI returns a clear 400
  • Subprocess RLIMIT_AS is a soft address-space cap, not a cgroup memory controller (Docker uses --memory)
  • Subprocess default-deny egress uses Linux unshare/bwrap when present; macOS stays credential-scrub only (network_isolated: false)
  • Single-node, no warm pool
  • TypeScript client is source-only (not published to npm)
S
Description
No description provided
Readme MIT
189 KiB
Languages
Python 95.9%
TypeScript 2.6%
Dockerfile 1.5%