#sandbox

Your agent is a security risk, so treat it like one. yoloAI does AI agent sandboxing right.

该仓库未发布描述。

Process isolation for Linux using namespaces, resource limits, cgroups, landlock and seccomp.

Kubernetes operator and aggregated apiserver implementing the agent-sandbox standard: warm microVM pools, ~33 ms claims through the Kubernetes API, and fleet scale with etcd off the claim path.

a13n - An open-source library and self-hosted platform for building and running your own agent systems—with managed agents, memory, sandboxes, computer use, and durable execution. Commercial Friendly. Built With Love 💜 @converge-ai-labs.

Embeddable Wasm sandbox for running untrusted mruby code from Ruby applications.

该仓库未发布描述。

Run multiple agents in parallel sandboxed VMs, with a single command, on your PC or in the cloud

AI agent sandbox

Public protocol source of truth for AgenC: Anchor program, zkVM guest, migrations, and generated protocol artifacts

Safe-by-default script sandbox in Go. Runs programs under default-deny OS isolation on Linux (bubblewrap + Landlock + seccomp): credential shielding, per-host network allowlisting via a host-side CONNECT proxy, resource limits, and a profile-then-grant consent flow. macOS support is planned.

BonkLM - LLM Security Guardrails with Interactive Setup Wizard

Rexec is an open-source platform that gives you instantly-available, network-isolated Linux terminals in the cloud, or lets you connect your own machines to a unified dashboard. Built for developers who need on-demand environments and secure remote access.

A Claude Skill to give your agent the ability to use a web browser

Local Python Sandbox

Kubernetes operator for stateful VM-backed pod workflows. Provides Hibernation CRD to suspend and wake VMs without pod deletion, and CocoonSet CRD to manage groups of related VM-backed pods with stable slot identities. Keeps stateful VM workloads inside native Kubernetes APIs.

a fast, rootless sandbox and virtual resource runtime. Run any workload in a real container, including an agent's tool-call or LLM-generated code.

A pure-Go WebAssembly sandbox for evaluating untrusted Jsonnet.

Kartikeya — standalone bwrap-sandboxed task queue + worker (the Willow fleet's Kart engine, extracted). Backend-agnostic: SQLite or Postgres.

Reusable detonation framework: run untrusted documents through disposable, hardened workers

Monkey-patched Python FS

Run AI coding agents in isolated containers with network filtering.

CLI coding tools in isolated Docker containers.

Lightweight microVMs for running anything.

Every Lua, everywhere — pure-Rust Lua 5.1-5.5, suite-passing, LuaRocks-compatible, wasm-ready.

The server-side runtime for AI agents. Durable & crash-safe (event-sourced), multi-worker scheduling, per-session sandboxing, human-in-the-loop, provider-neutral. Host agents like Claude Code on your own infra — self-hostable, no vendor lock-in.

该仓库未发布描述。

Give agents an operating system as a library. Runs in your existing backend – no sandboxes, VMs, or SaaS. Powered by WebAssembly & V8 isolates.

Cloud-agnostic Python SDK for extensible cloud service integrations

该仓库未发布描述。