Openshell Deepagent

OpenShell Deep Agent

A general-purpose coding agent that runs inside an [NVIDIA OpenShell](https://github.com/NVIDIA/OpenShell) sandbox, orchestrated by [Deep Agents](https://docs.langchain.com/oss/python/deepagents/overview) and powered by NVIDIA Nemotron. The agent writes and executes code in an isolated, policy-governed Linux environment — no cloud dependency required.

What is OpenShell?

[OpenShell](https://github.com/NVIDIA/OpenShell) is like a browser's security model, but for agent code execution. It's an on-prem sandbox that lets agents write and run code while enforcing **policies** that control filesystem access, network access, and process permissions. Think of it as a more secure Docker container — the agent can self-evolve and learn new skills, but it can't be tricked into leaking data or running destructive commands.

This matters for LangChain because deep agents connect to real data sources (Linear, Slack, Salesforce). You want the agent to learn new things on the fly, but you don't want to rely on prompt instructions alone to prevent misuse — something external to the agent has to enforce security, which is what OpenShell does.

What are Deep Agents?

The easiest way to start building agents and applications powered by LLMs — with built-in capabilities for task planning, file systems for context management, subagent-spawning, and long-term memory. You can use deep agents for any task, including complex, multi-step tasks.

We think of deepagents as an "agent harness". It is the same core tool calling loop as other agent frameworks, but with built-in tools and capabilities.

deepagents is a standalone library built on top of LangChain's core building blocks for agents. It uses the LangGraph runtime for durable execution, streaming, human-in-the-loop, and other features.

Architecture

┌─────────────────────────────────────────────────────┐
│  LangGraph Dev Server (http://127.0.0.1:2024)       │
│