AgentToolDrop

12 of 43

smolvm

Runs agent workloads in local microVMs with their own kernel, which you can checkpoint, branch and pack

Install with Homebrew or the install script

brew install smol-machines/tap/smolvm                  # macOS (Homebrew)
curl -sSL https://smolmachines.com/install.sh | bash   # macOS + Linux

Copied from the project's docs at github.com/smol-machines/smolvm

Why you'd want it

smolvm boots any OCI image as a hardware-virtualized VM with its own guest kernel, using Hypervisor.framework on macOS, KVM on Linux or the Windows Hypervisor Platform, with no Docker daemon. Networking is off by default, egress can be limited to named hosts, and credential substitution lets code use a secret without reading it. A running machine can be checkpointed, rewound or branched into copy-on-write copies, packed into a single .smolmachine file, and driven from Node, Python or Rust SDKs that run in your process without a daemon.

Best for

  • Running untrusted agent code on your own machine
  • Branching an agent's environment mid-run
  • Embedding VMs in your own app

Use it when

You want each agent task in a real VM on your laptop or server, with egress allow lists and the option to snapshot or branch the machine.

Skip it if

Containers or a hosted sandbox already give you enough isolation. The README calls safety a shared responsibility: every folder you mount, port you open, host you allow and secret you forward becomes usable by the workload. GPU acceleration is unavailable on Windows, and Intel macOS is marked untested.

Editor's note

The README says smolvm is the open-source runtime behind smol machines and is not affiliated with other projects named SmolVM.

Related tools

Same category or shared tags.

  1. Rank 14

    nono

    Sandbox

    Runs coding agents in a least-privilege sandbox with per-tool filesystem, network and credential rules

    • score 47
    • 4.5k stars
    • 17.5 stars/day since creation
    • pushed
    • v0.80.0 on
    • Apache-2.0

    Install with Homebrew

    brew install nono
    Use it when, skip it if, stats for nono

    Why you'd want it

    nono wraps an agent such as Claude Code, Codex or Pi in a sandbox with no daemon, container or VM, and shows the exact filesystem, network and credential access before applying it. Profiles for common agents are pulled from the nono registry and can be extended into an editable JSON profile. Tools the agent calls, such as git, gh, curl or kubectl, can run in their own command sandboxes with separate filesystem and network policy, and credentials reach them through a proxy that can filter by HTTP method and path.

    Best for

    • Sandboxing Claude Code, Codex or Pi locally
    • Scoping credentials to one tool
    • Sharing reviewed agent profiles with a team

    Use it when

    You want to run a coding agent on your own machine with explicit filesystem, network and credential limits, without setting up containers or VMs.

    Skip it if

    You want VM or container isolation; the README says nono runs with no daemon, container or VM. APIs may still change before 1.0, and Windows is supported through WSL2. It runs agent commands and brokers credentials, so review a profile's grants before pulling it from the registry.

    Stats

    Stars
    4,504
    Last push
    Release
    v0.80.0,
    License
    Apache-2.0
    Pricing
    Open source
    Works with
    Claude Code, Codex, macOS, Linux, Windows
    Stars per day, since repo creation
    17.5
    New and rising score
    47 of 100

    Full stat block for nono

  2. Rank 43

    Container Use

    Sandbox

    Gives each coding agent its own container and git branch so several can work in parallel without conflicts

    • score 11
    • 4.1k stars
    • 8.0 stars/day since creation
    • pushed
    • v0.4.2 on
    • Apache-2.0

    Install with Homebrew and add to Claude Code

    brew install dagger/tap/container-use
cd /path/to/repository
claude mcp add container-use -- container-use stdio
    Use it when, skip it if, stats for Container Use

    Why you'd want it

    Container Use is an MCP server, run from a CLI and built on Dagger, that gives each agent a fresh container on its own git branch, so you can run several agents at once and discard failed attempts. You can see the full command history and logs of what an agent did, drop into its terminal to take over, and review its work with a normal git checkout of its branch. Setup is the same for any MCP-compatible agent, adding container-use stdio as an MCP server, and the docs have guides for Claude Code, Cursor, Windsurf, Zed, Codex, Gemini CLI and others.

    Best for

    • Running several coding agents in parallel
    • Keeping agent changes off your working tree until reviewed
    • Claude Code, Cursor and other MCP agents

    Use it when

    You want coding agents to work in isolated, reviewable environments on your own machine instead of directly in your checkout.

    Skip it if

    You do not have Docker and Git installed, which the quickstart requires. The project describes itself as experimental and in early development.

    Stats

    Stars
    4,070
    Last push
    Release
    v0.4.2,
    License
    Apache-2.0
    Pricing
    Open source
    Works with
    Claude Code, Cursor, Windsurf, Zed, Codex, Gemini CLI, VS Code, Any MCP client
    Stars per day, since repo creation
    8.0
    New and rising score
    11 of 100

    Full stat block for Container Use

  3. Qwen Code

    Coding agent Standard

    Runs a coding agent in your terminal that works with Qwen, OpenAI, Anthropic, Gemini or local models

    • 28.5k stars
    • 59.8 stars/day since creation
    • pushed
    • sdk-typescript-v0.1.19 on
    • Apache-2.0

    Install with npm (Node.js 22+)

    npm install -g @qwen-code/qwen-code@latest
    Use it when, skip it if, stats for Qwen Code

    Why you'd want it

    Qwen Code was originally based on Gemini CLI v0.8.2 and has been developed independently since v0.1. It supports OpenAI, Anthropic, Gemini and Qwen APIs, other third-party providers and local models through Ollama or vLLM, and you can switch provider at runtime. Beyond the terminal it has VS Code, Zed and JetBrains integrations, a desktop app, an experimental web UI, a headless qwen -p mode and TypeScript, Python and Java SDKs.

    Best for

    • Running Qwen models in an agent
    • Mixing hosted and local models
    • Scripted runs in headless mode

    Use it when

    You want a terminal coding agent that is not tied to one model vendor, or you specifically want to use Qwen models.

    Skip it if

    You were counting on the free Qwen OAuth tier, which the docs say was discontinued on 2026-04-15; you now need an API key, an Alibaba Cloud plan or a local model. Its docs say optional usage statistics can be collected, so check the setting.

    Stats

    Stars
    28,487
    Last push
    Release
    sdk-typescript-v0.1.19,
    License
    Apache-2.0
    Pricing
    Open source
    Works with
    macOS, Linux, Windows, VS Code, JetBrains, Zed, Web
    Stars per day, since repo creation
    59.8
    Tier
    Standard, not ranked

    Full stat block for Qwen Code

For agents

The same entry, as files an agent can read without scraping this page. 

# read one entry
curl -s https://agenttooldrop.com/tools/smolvm.md

# list the MCP servers in the catalog
curl -s https://agenttooldrop.com/tools.json \
  | jq -r '.tools[] | select(.category=="mcp-server") | .slug'