Docker or E2B — which should I use?

Docker keeps everything on your machines at no extra cost. E2B offloads isolation to managed cloud microVMs, which suits hosted deployments.

Is the sandbox on by default?

No. Enable it in Admin → Tools → Runtime or with FASTCLAW_SANDBOX_ENABLED=true.

Integration · Runtime

Use Docker as FastClaw's sandbox backend to isolate agent commands on your own hardware — or deploy FastClaw itself with Docker Compose.

Agent exec and file tools run inside a container on your host, with skills and workspace hydrated automatically.

Pick any image with the tools your agents need, such as python:3.12-slim, or build the bundled sandbox image.

The repo ships a docker-compose.yml that runs FastClaw with Postgres — bootstrap is env-only, no config file to mount.

Give agents a safe place to run code and shell commands without touching your host — see the sandbox feature.
Expose code-running agents to a community while keeping the server locked down.
Need managed isolation instead? Use E2B.

Make sure Docker is running on the machine that runs FastClaw.
In Admin → Tools → Runtime, enable the sandbox, choose Docker and set the image.
Or use environment variables:

FASTCLAW_SANDBOX_ENABLED=true
FASTCLAW_SANDBOX_BACKEND=docker
FASTCLAW_SANDBOX_IMAGE=python:3.12-slim

To deploy FastClaw itself in containers, start from deploy/docker/docker-compose.yml in the FastClaw repo, which pairs it with Postgres.

Set FASTCLAW_SANDBOX_ENFORCE=1 on a self-hosted install to lock every exec and file call inside the sandbox.

Run every FastClaw agent command in an isolated E2B cloud sandbox — skills and workspace hydrated in, files synced back after each call.

Swap FastClaw's default SQLite file for PostgreSQL to run multiple pods behind a load balancer, with automatic schema migrations.

Give FastClaw agents external tools through the Model Context Protocol — add http or stdio MCP servers per agent from the console.

Start free in FastClaw Cloud, or self-host the single binary in a minute.