# Sandcastle — AI Coding Tool Review & Pricing

> Sandcastle is a TypeScript library and CLI from Matt Pocock's AI Hero for orchestrating AI coding agents in isolated sandboxes. A single `sandcastle.run()` call starts an agent on its own branch, sandboxes it with a chosen provider, and merges the commits back, making it easy to run parallel AFK agents and review pipelines.

Source: https://ai.dosa.dev/tools/sandcastle · Updated Nov 3, 2026 · Category: Autonomous & Async Agents · Published by dosa.dev

## At a glance

- **Tool**: Sandcastle
- **Company**: AI Hero (Matt Pocock)
- **Category**: [Autonomous & Async Agents](https://ai.dosa.dev/category/async-agents)
- **Pricing model**: Open Source
- **Official website**: https://www.aihero.dev/sandcastle

## Key features

- `sandcastle.run()` API that sandboxes an agent, applies a branch strategy and merges commits back
- Agent providers for Claude Code, Codex, Copilot, Cursor, OpenCode and Pi
- Native Bubblewrap and Seatbelt sandboxes plus Docker, Podman, E2B, Modal, Daytona, Vercel and Sprites providers
- `sandcastle init` templates for simple loops, parallel planners and review pipelines
- Hooks, iterations, structured output, agent session resume and run logs

## Pricing

**Model**: Open Source

Sandcastle is free and open-source under the MIT license, published on npm as `@ai-hero/sandcastle`. You pay separately for the coding agents and models you run and for any cloud sandbox provider (E2B, Modal, Daytona, Vercel or Fly.io Sprites) you choose.

## Who it is for

- **Best for**: TypeScript developers who want to script parallel, unattended coding-agent runs with real isolation and git-based merge-back in their own repos or CI.
- **Not ideal for**: Developers looking for a GUI agent manager or a hosted agent service, since Sandcastle is a pre-1.0 library that requires writing orchestration code.

## Verdict

A well-designed, provider-agnostic building block for running many sandboxed coding agents in parallel, backed by an active maintainer and fast-growing adoption.

## Recent updates

v0.13.0 (November 2, 2026) made native Bubblewrap and Seatbelt sandboxes the documented default and extended session resume to more agents; v0.14.0 the same day added an E2B isolated sandbox provider. The repository passed 8,800 GitHub stars, appeared on GitHub's weekly trending list, and the npm package reports about 234,000 downloads a week.

## Frequently asked questions

### Is Sandcastle free?

Yes - Sandcastle is Open Source. Sandcastle is free and open-source under the MIT license, published on npm as `@ai-hero/sandcastle`. You pay separately for the coding agents and models you run and for any cloud sandbox provider (E2B, Modal, Daytona, Vercel or Fly.io Sprites) you choose.

### Is Sandcastle open source?

Yes - Sandcastle is open source. Sandcastle is free and open-source under the MIT license, published on npm as `@ai-hero/sandcastle`. You pay separately for the coding agents and models you run and for any cloud sandbox provider (E2B, Modal, Daytona, Vercel or Fly.io Sprites) you choose.

### Who is Sandcastle best for?

TypeScript developers who want to script parallel, unattended coding-agent runs with real isolation and git-based merge-back in their own repos or CI.

### Who is Sandcastle not ideal for?

Developers looking for a GUI agent manager or a hosted agent service, since Sandcastle is a pre-1.0 library that requires writing orchestration code.

### What are the best Sandcastle alternatives?

The closest Sandcastle alternatives on ai.dosa.dev are Jules by Google, OpenHands, SWE-agent - all listed under Autonomous & Async Agents.

### Who makes Sandcastle?

Sandcastle is developed by AI Hero (Matt Pocock). It is listed in the Autonomous & Async Agents category on ai.dosa.dev.

## Top alternatives

- [Jules by Google](https://ai.dosa.dev/tools/jules) (Freemium) — Developers who want to offload repetitive, non-urgent tasks like dependency version bumps, lint fixes, test scaffolding, and resolving CI failures in GitHub-hosted projects.
- [OpenHands](https://ai.dosa.dev/tools/openhands) (Freemium) — Individual developers and software teams looking to automate repetitive coding tasks, repository maintenance, and issue resolution through autonomous agents.
- [SWE-agent](https://ai.dosa.dev/tools/swe-agent) (Open Source) — Autonomous resolution of software engineering issues in real GitHub repositories and performing custom developer tasks using language models.
- [mini-swe-agent](https://ai.dosa.dev/tools/mini-swe-agent) (Open Source) — Researchers and engineers who want a small, transparent coding agent baseline for benchmarks, fine-tuning or sandboxed batch runs, and developers who want a simple bash-only agent in the terminal.
- [Sweep](https://ai.dosa.dev/tools/sweep) (Freemium) — Individual developers and teams working primarily in JetBrains IDEs who want a native-feeling assistant for routine coding tasks, bug fixes, and autocompletions.
- [Ellipsis](https://ai.dosa.dev/tools/ellipsis) (Freemium) — Teams looking to centralize and govern AI coding agents across repositories with secure, non-personal credentials and persistent cloud-based execution.
- Full alternatives guide: https://ai.dosa.dev/tools/sandcastle/alternatives