Agent frameworks & infrastructure
Agent sandboxes & runtimes
1,721 agents · most starred first
Agent sandboxes & runtimes, page 17 of 18
A local-first, auditable job runner for AI-agent automation
Agent Environment public package coordinate scaffold
Agent Environment public package coordinate scaffold
DK Agent OS — Autonomous Developer Runtime & Commercial SDK Platform
Haze -- pool your own machines into a private compute network
An event-sourced runtime for durable, crash-resumable LLM agents
Autonomous agent for small, lawful multi-runtime code execution (Go, Python, Node.js, Rust), Go/Linux automation, and research microtasks, paid in BTC on-chain.
Coder workspace sandbox provider for the Vercel AI SDK v7 HarnessAgent
`HarnessV1` adapter backed by an NPM-installed [Agent Client Protocol](https://agentclientprotocol.com/) version 1 implementation. The adapter ships a bridge process that runs inside a sandbox and talks to the host over a WebSocket on a sandbox-proxied lo
Run AI coding agents in isolated Docker or Podman containers
Deterministic onboarding and maintenance for hardened coding-agent containers
opsen as an agentsh secure-sandbox adapter — machine, model and tool spend on one bill.
AgentShip runtime service — the REST/SSE/WS surface over an agent, with pluggable auth adapters and tenant isolation. The irreducible in-app core a gateway cannot replace.
The s3:// storage backend for dirigent, compatible with any S3 API.
Linux runtime bridge between FastLLM WebUI and Pi Agent
Local Unified Harness runtime and native bindings for Vibe
Run OpenAI Agents SDK sandboxes on opsen - machine, model and tool spend on one bill.
Run Pydantic AI agents on opsen — machine, model and tool spend on one bill.
agent tasks for Render Workflows
Run any AI agent on a FlockTab tab: tab claude, tab codex, tab <command>.
Autonomous AI Agent Platform & Control Plane
Mobius CLI for pairing and running local AI agent services.
Terminal UI for managing multiple AI agent CLIs
Anantham V2 — Programmable AI Agent Operating Environment
User interface for AI agent assistants that run Python code safely on your device through an intuitive chat interface. Supports local and cloud LLM models with built-in security features.
Universal sandbox contract, policy algebra, and outcome classification for ai-agent-sdk
Cloud AI agent runtime
Your model writes the code. This runs it where it can't touch your machine: a rootless Linux container, no daemon, no VM, no cloud, no account. A kernel boundary, not a microVM: for deliberately hostile code, use one.
Run AI coding agents (Claude Code, OpenAI Codex, GitHub Copilot CLI) inside Docker against a dedicated Git worktree.
DeepSeek Harness native macOS desktop app, background daemon, and autonomous agent environment manager
Run AI coding agents in isolated sandboxes
A lightweight bwrap sandbox for AI agents
Reconnectable native agent workers with durable input receipts and events
Self-hosted code execution sandbox for AI agents
Placeholder package reserving the agentos-execution-kernel name on PyPI. Real implementation tracked in docs/.
Placeholder package reserving the agentos-execution-kernel-client name on PyPI. Real implementation tracked in docs/.
The reliability layer for AI agents with write access: policy-first, code-gated, read-back-verified multi-app execution
Python SDK for the camelAI hosted agent runtime
Sandboxed Claude Code, Codex and Pi: a bubblewrap jail with an egress allowlist, via uvx
Docker Sandboxes (sbx) microVM sandbox backend for Deep Agents
Direct and durable production execution for HarborRAG.
Compile a plain-text requirement into a reusable, verified, sandboxed function
Async Azure Container Apps Sandbox backend for LangChain Deep Agents
Runtime sandbox integration for Deep Agents
A maf-sandbox router as a Deep Agents sandbox backend — any backend the router admits (Docker, ACA Sandboxes) behind LangChain's BaseSandbox, with the router's isolation floor and egress policy still enforced.
OpenAI Agents SDK adapter for MemSandbox
Webhook-managed OpenAI Agents API sessions in Modal sandboxes
Configurable local runner for Papaya agent-client jobs
A small, inspectable runtime for durable agent tool execution, recovery, cancellation, and evidence.
PyO3 bindings for the Rust CoreLoop.
Agent hosting for the ygo74 agent runtime: endpoints, discovery, sessions and human approval
agent-exec review: compare features, pricing, use cases, access model, and alternatives for this Sandboxes agent in 2026. Pay-per-use code execution for AI agents. HTTP 402, no account, no API key.
Node process-confinement backends and in-process path fence for ai-agent-sdk
Persistent Linux workspaces for agentic development.
Run coding agents in rootless Podman workspace boxes
Agent Sandbox SDK — Create, manage, and interact with cloud sandboxes for AI agents
Agent Sandbox Server SDK — container-side runtime service for Agent Sandbox templates
Policy-driven process isolation and supervision for agent runtimes
Manage an isolated Xubuntu development VM for coding agents
Daytona sandbox provider for Scale AI's AgentEnv framework
SDK and CLI for building, deploying and running agent environments and the tasks that grade agents in them.
Open data-plane protocol and server SDK for agent environments.
The Agentic OS Runner: an activity-only Temporal executor for workspaces, Agent Runtimes and verb seams
Bounded multi-runtime agent session pools for Python
Axon — 单用户 CLI Agent Harness:沙箱工具网关、双模型接入、可续跑与项目记忆
Token-efficient remote execution, AST outlining, and safe patching tool for AI agents.
The Node.js runtime the local Comfy agent's document host runs on.
Daytona sandbox drivers for the Anthropic SDK's computer and browser toolsets
E2B browser and computer drivers for Anthropic toolsets
Node execution agent for Fadenstack cluster-managed runtimes.
Harakiri sandbox backends for Python Deep Agents
Tensorlake sandbox and Cloud Volume filesystem backends for Deep Agents
Lizard Sandboxes integration for LangChain Deep Agents
Promigence sandbox backend for LangChain DeepAgents
Submilli code mode for LangChain: run an agent's tool calls as TypeScript programs on submilli-server, under a Blueprint's policy.
Python SDK and CLI for the Mirasoth AgentSpace API - isolated cloud workspaces for agent execution
Sandboxed Python code execution tool for Haystack Agents, powered by Pydantic Monty
Placeholder package for MouseCat AI.
Drop-in run() replacement and sandbox tools that give openai-agents agents an isolated microsandbox workspace.
Runtime Cloud sandboxes for the OpenAI Agents SDK: run a SandboxAgent in a Firecracker microVM
A security-first, zero-dependency agent runtime with an execution-based evaluation harness.
Praxis — a trusted-local workspace agent runtime for files, code, data, and private knowledge.
Run Claude Managed Agents self-hosted environment workers on Promigence sandboxes
Run mini-swe-agent (SWE-bench) instances on Promigence sandboxes
Promigence sandbox client for the OpenAI Agents SDK
Placeholder for Python bindings to agentplane, a durable, replayable agent runtime.
Pydantic AI workspaces in Lizard Sandboxes.
Self-hostable disposable sandboxes for AI agents, with snapshot, rollback, and fork.
Run the code of smolagents' CodeAgent in a local Hyperlight micro-VM
Run smolagents Python code in Lizard Sandboxes.
Terminal and web AI agent that answers requests by driving an LLM in a tool-call loop inside a sandboxed container pod.
Runtime sandboxes as a Pydantic AI toolset: shell, files and listings in a Firecracker microVM
Runtime Cloud sandboxes as a smolagents code executor: a CodeAgent's Python runs in a Firecracker microVM
Deep Agents sandbox backend for the Harakiri control plane.
Agent37 sandbox backend for LangChain Deep Agents
Prized Sandboxes (Firecracker microVMs) as a sandbox backend for LangChain Deep Agents
Prized Sandboxes (Linux microVMs) as a sandbox client for the OpenAI Agents SDK
Run the OpenAI Agents SDK's sandbox commands in local Hyperlight micro-VMs
Official Python SDK for QuantaHost object storage (QuantaStore). Uploads go direct to storage via presigned URLs; zero dependencies.
Tool definitions that let an LLM agent work inside a ClientVM instance.