← all releases
Anthropic·Oct 23, 2026·v0.0.81·open source
Anthropic Sandbox Runtime (srt)
The Anthropic Sandbox Runtime is a lightweight, open-source tool designed to enforce filesystem and network restrictions on arbitrary processes at the OS level. Developed primarily for Claude Code, it provides a secure environment for running AI agents, local MCP servers, and shell commands without the overhead of traditional containers.
AI AgentsAgentic AIInfrastructureTesting
overview
The Anthropic Sandbox Runtime is a lightweight, open-source tool designed to enforce filesystem and network restrictions on arbitrary processes at the OS level. Developed primarily for Claude Code, it provides a secure environment for running AI agents, local MCP servers, and shell commands without the overhead of traditional containers.
key features
- 01OS-level process isolation without container overhead
- 02Granular filesystem and network access control
- 03Support for sandboxing AI agents, MCP servers, and bash commands
- 04Configurable security boundaries for arbitrary processes
related productsbrowse all →
AIO Sandboxopen sourceAIO Sandbox is an integrated environment designed for AI agents, combining a browser, terminal, filesystem, VSCode, Jupyter, and MCP Server into a single Docker container. This unified setup allows seamless development and execution of AI agents without the need for multiple services.Fractalsopen sourceFractals is an open-source recursive agentic task orchestrator developed by TinyAGI. It enables the decomposition of high-level tasks into self-similar trees of executable subtasks, which are then executed by an agent swarm in isolated git worktrees.MiroThinkeropen sourceMiroThinker is an open-source search agent model developed by MiroMindAI, designed for tool-augmented reasoning and real-world information seeking. It aims to match the deep research capabilities of leading AI models like OpenAI's Deep Research and Google's Gemini Deep Research.