Havuz

The most comprehensive comparison of platforms offering cloud runtime / sandbox / VM compute for agents. By boat.dev.

🔒 Jail OpenCode, Claude Code, Copilot CLI and OmO in micro-VMs: only your project folders, only the networks you allow.

A new kind of IDE for agentic engineering

Run untrusted Erlang inside an Elixir host, on the same BEAM, with no ambient authority.

Give any AI agent its own disposable Linux machine. Firecracker microVM sandboxing with millisecond boot times, full network access, and nat

A lightweight and security-focused Python implementation of OpenClaw.

Mobile-native agent SDK for building on-device AI agents with sandboxed execution, extensible tools, and x-app / x-agent / x-channel connect

Detonate files & URLs in cloud malware sandboxes (Hybrid Analysis, tria.ge, ANY.RUN) and enrich IOCs across MalwareBazaar, ThreatFox, URLhau

Self-hosted, privacy-first Markdown editor & live preview. Documents never leave your browser: no backend, no tracking, strict CSP, sandboxe

Claude Code delegates coding to isolated CLI agents it doesn't trust (Codex, OpenCode, Pi, Pythinker), freezes what they produce, verifies i

Run a command in a sandbox with the same flags on every OS — one capability model compiled to Seatbelt (macOS), gVisor (Linux), and AppConta

CLI for Local Isolated Environments for Coding Agents

Execute agent-generated TypeScript in isolated WebAssembly, with runtime-enforced semantic permissions.

Agents you can inspect, replay and trust. An agent framework for TypeScript and Python, built on an append-only event log.

Agent dashboard runs reliability scans, skills, and network access probes.

Vyomi — the Multi-Cloud Digital Twin Runtime. A faithful, running replica of AWS, GCP & Azure on your machine (real backends, standard SDKs)

Self-hosted platform for running AI coding agents (Claude Code, Codex, …) in isolated sandboxes – web UI, CLI, API and MCP

A fast, embeddable JavaScript engine for Go: QuickJS-NG reimplemented in pure Go, no cgo.

Governed AI agents for sensitive data and business-critical apps. Agents never hold credentials: a gateway checks every call against per-res

A sandbox service for running untrusted, agent-authored code that reports the isolation boundary each run executed behind. Lessons, no clone

在網路受限的 Docker Dev Container 裡安全執行 Claude Code，透過 iptables/ipset 防火牆白名單限制 AI Agent 的對外連線，降低資料外洩與誤操作風險。改自 Anthropic 官方 .devcontainer 參考實作。

OpenShell Kubernetes Operator

基于 Rust、Slint 和 Skia 的 AI 算法可视化桌面应用，支持代码对话、文件附件、隔离验证与交互式动画演示。

Open-source continuation of Daytona — secure, elastic sandbox infrastructure for AI code execution and agent workflows. Community fork of th

Prompt-injection, jailbreak and secret-leak defense for DeepSeek Harness: Aho-Corasick detection with allow/ask/block interception and sanit

Run OpenClaw, Hermes and your own agents in microsandbox microVMs. One reviewable TOML role sets image, allowed domains and host-bound secre

Ocean Drive, Miami Beach at golden-hour sunrise: a photoreal open-world Three.js sandbox in the browser. Walk, drive any car, hit the beach.

Your dashboard is data. Any AI can build it; any human can edit it. A protocol + runtime for agent-composable dashboards — layout-as-data, o

Polyglot, deterministic, sandboxed WASM runtime. Run or compile any language to one sealed, reproducible artifact.

Autonomous coding-agent loop. Runs Claude Code or Codex against your repo in a Docker sandbox, iterating implementer → reviewer until the pl

epoint.az ödəniş gateway-inin lokal sandbox-u. Bir konteynerdə: API, ödəniş səhifəsi, callback-lər və dashboard.

Bounded repository context and verified patches for Codex and Claude Code. Explicit edit grants, source fingerprints, isolated worktrees, ac

Run a coding agent in a staged workspace, review its file changes, and apply only what you keep.

Run Claude Code (or another coding agent) unattended, inside a disposable microVM, with its work committed to git so you can review everythi

A hidden Omarchy desktop for your AI agents: they launch, drive and screenshot apps and shell plugins there, never on yours.

🪝 透明可控的 AI Agent 终端 | 状态机 · 多层记忆系统 · 沙箱防御 · 完整审计 · 多人设多技能

MCP server: build → see → interact → debug loop over native GUI apps

Security-first, self-hosted AI agents - isolation you can prove, not just promise.

Train computer-use agents end-to-end on real desktops, at scale: a high-performance runtime where 8K+ live environments on one laptop thread

Run any AI agent — Claude Code, GAP, LangGraph DeepAgents — in a hermetic sandbox. Reference implementation of the Harness Protocol.

Self-hosted sandboxes for AI agents: Docker, Firecracker or Apple Virtualization microVMs. Your real API keys never enter the sandbox. Part

炼妖壶 — run Claude Code, Codex, or any coding agent in a deny-default sandbox with VPN-only network egress. sandbox-exec + PF on macOS, Landlo

Self-hosted AI gateway with a built-in dashboard. OpenAI-compatible API, multi-tenant, sandboxed code execution, MCP integrations, file stor

多文件夹 workspace：让 DSH（DeepSeek Harness）的 Agent 不只能读写主目录，还能同时读写你添加的其他文件夹。Multi-folder workspace for DeepSeek Harness: let the agent read and w

Give Pi one typed tool instead of a toolbox.

The runtime substrate for agentic systems — one Go binary, six LLM providers, MCP-native, configurable as a managed sandbox or full agentic

One isolated Lima VM per GitHub project — sandboxed Claude Code/Codex, Docker, Node, signed git

One-click sandboxes for AI agents, with egress policy, MCPs, skills and audit. Built on NVIDIA OpenShell.

Vmware Workstation 17 Pro - Local Windows build featuring every professional module and a simple install routine.

Run Claude Code, Codex and other coding agents in a local VM on a clone of your project, then review their changes like a pull request. Mac