I set up my first Docker SBX kit, and here's how I did it

Description and use

A reusable SBX kit that automatically starts an MLflow server in a Docker sandbox, eliminating manual setup and errors.

Problem

AI engineers repeatedly configure sandbox environments manually, leading to configuration drift and inconsistent setups.

Outcome

The mixin kit enables instant, reproducible MLflow-enabled sandboxes, eliminating manual setup and reducing errors.

Workflow

• Plan architecture and sketch components
• Create folder structure with spec.yaml, Dockerfile, CLAUDE.md, scripts, README
• Write Dockerfile to include heavy dependencies
• Write CLAUDE.md to expose MLflow capabilities to Claude Code
• Write spec.yaml to define mixin, environment variables, network rules, install commands, and startup script
• Debug and iterate on spec.yaml until it works
• Publish kit for reuse

Evidence

A Docker Sandbox Kit (SBX Kit) solves this. It's a declarative YAML file (spec.yaml) that configures your sandbox environment automatically at creation time.

Source

The goal: A mixin kit that automatically starts an MLflow tracking server when the sandbox is created, so Claude Code can log ML experiments, track metrics, and version models right out of the box.

Source