[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-3736":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":25,"hasPages":25,"topics":26,"createdAt":10,"pushedAt":10,"updatedAt":27,"readmeContent":28,"aiSummary":29,"trendingCount":16,"starSnapshotCount":16,"syncStatus":30,"lastSyncTime":31,"discoverSource":32},3736,"NemoClaw","NVIDIA\u002FNemoClaw","NVIDIA","Run agents like Hermes and OpenClaw more securely inside NVIDIA OpenShell with managed inference","https:\u002F\u002Fdocs.nvidia.com\u002Fnemoclaw\u002Flatest\u002F",null,"TypeScript",21126,2811,114,274,0,19,198,823,96,45,"Apache License 2.0",false,"main",true,[],"2026-06-12 02:00:53","\u003C!--\n  SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.\n  SPDX-License-Identifier: Apache-2.0\n-->\n\n# 🦞 NVIDIA NemoClaw: Reference Stack for Running OpenClaw in OpenShell\n\n\u003C!-- start-badges -->\n[![License](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLicense-Apache_2.0-blue)](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FNemoClaw\u002Fblob\u002Fmain\u002FLICENSE)\n[![Security Policy](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FSecurity-Report%20a%20Vulnerability-red)](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FNemoClaw\u002Fblob\u002Fmain\u002FSECURITY.md)\n[![Project Status](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fstatus-alpha-orange)](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FNemoClaw\u002Fblob\u002Fmain\u002Fdocs\u002Fabout\u002Frelease-notes.md)\n[![Discord](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FDiscord-Join-7289da)](https:\u002F\u002Fdiscord.gg\u002FXFpfPv9Uvx)\n\u003C!-- end-badges -->\n\n\u003C!-- start-intro -->\nNVIDIA NemoClaw is an open source reference stack that simplifies running [OpenClaw](https:\u002F\u002Fopenclaw.ai) always-on assistants more safely.\nIt installs the [NVIDIA OpenShell](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FOpenShell) runtime, part of NVIDIA Agent Toolkit, which provides additional security for running autonomous agents.\n\u003C!-- end-intro -->\n\n> **Alpha software**\n>\n> NemoClaw is available in early preview starting March 16, 2026.\n> This software is not production-ready.\n> Interfaces, APIs, and behavior may change without notice as we iterate on the design.\n> The project is shared to gather feedback and enable early experimentation.\n> We welcome issues and discussion from the community while the project evolves.\n\nNemoClaw adds guided onboarding, a hardened blueprint, state management, OpenShell-managed channel messaging, routed inference, and layered protection on top of the [NVIDIA OpenShell](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FOpenShell) runtime. For the full feature list, refer to [Overview](https:\u002F\u002Fdocs.nvidia.com\u002Fnemoclaw\u002Flatest\u002Fabout\u002Foverview.html). For the system diagram, component model, and blueprint lifecycle, refer to [How It Works](https:\u002F\u002Fdocs.nvidia.com\u002Fnemoclaw\u002Flatest\u002Fabout\u002Fhow-it-works.html) and [Architecture](https:\u002F\u002Fdocs.nvidia.com\u002Fnemoclaw\u002Flatest\u002Freference\u002Farchitecture.html).\n\n## Getting Started\n\nFollow these steps to install NemoClaw and run your first sandboxed OpenClaw agent.\n\n\u003C!-- start-quickstart-guide -->\n\n### Prerequisites\n\nBefore getting started, check the prerequisites to ensure you have the necessary software and hardware to run NemoClaw.\n\n#### Hardware\n\n| Resource | Minimum        | Recommended      |\n|----------|----------------|------------------|\n| CPU      | 4 vCPU         | 4+ vCPU          |\n| RAM      | 8 GB           | 16 GB            |\n| Disk     | 20 GB free     | 40 GB free       |\n\nThe sandbox image is approximately 2.4 GB compressed. During image push, the Docker daemon, k3s, and the OpenShell gateway run alongside the export pipeline, which buffers decompressed layers in memory. On machines with less than 8 GB of RAM, this combined usage can trigger the OOM killer. If you cannot add memory, configuring at least 8 GB of swap can work around the issue at the cost of slower performance.\n\n#### Software\n\n| Dependency | Version                          |\n|------------|----------------------------------|\n| Node.js    | 22.16 or later |\n| npm        | 10 or later |\n| Platform   | See below |\n\n#### OpenShell Lifecycle\n\nFor NemoClaw-managed environments, use `nemoclaw onboard` when you need to create or recreate the OpenShell gateway or sandbox.\nAvoid `openshell self-update`, `npm update -g openshell`, `openshell gateway start --recreate`, or `openshell sandbox create` directly unless you intend to manage OpenShell separately and then rerun `nemoclaw onboard`.\n\n#### Container Runtimes\n\nThe following table lists tested platform and runtime combinations.\nAvailability is not limited to these entries, but untested configurations may have issues.\n\n\u003C!-- platform-matrix:begin -->\n| OS | Container runtime | Status | Notes |\n|----|-------------------|--------|-------|\n| Linux | Docker | Tested | Primary tested path. |\n| macOS (Apple Silicon) | Colima, Docker Desktop | Tested with limitations | Install Xcode Command Line Tools (`xcode-select --install`) and start the runtime before running the installer. |\n| DGX Spark | Docker | Tested | Use the standard installer and `nemoclaw onboard`. For an end-to-end walkthrough with local Ollama inference, see the [NVIDIA Spark playbook](https:\u002F\u002Fbuild.nvidia.com\u002Fspark\u002Fnemoclaw). |\n| Windows WSL2 | Docker Desktop (WSL backend) | Tested with limitations | Requires WSL2 with Docker Desktop backend. |\n\u003C!-- platform-matrix:end -->\n\nFor platform-specific pre-setup (for example, Windows WSL 2), see [Prerequisites](https:\u002F\u002Fdocs.nvidia.com\u002Fnemoclaw\u002Flatest\u002Fget-started\u002Fprerequisites.html).\n\n### Install NemoClaw and Onboard OpenClaw Agent\n\nDownload and run the installer script.\nThe script installs Node.js if it is not already present, then runs the guided onboard wizard to create a sandbox, configure inference, and apply security policies.\n\n> **ℹ️ Note**\n>\n> NemoClaw creates a fresh OpenClaw instance inside the sandbox during the onboarding process.\n>\n> The installer runs as your normal user and does not require `sudo` or root.\n> It installs Node.js via nvm and NemoClaw via npm, both into user-local directories.\n> Docker must be installed and running before you run the installer. Installing Docker may require elevated privileges on Linux.\n\n```bash\ncurl -fsSL https:\u002F\u002Fwww.nvidia.com\u002Fnemoclaw.sh | bash\n```\n\nThe piped installer prompts through your terminal. In headless scripts or CI,\npass explicit acceptance to the `bash` side of the pipe:\n\n```bash\ncurl -fsSL https:\u002F\u002Fwww.nvidia.com\u002Fnemoclaw.sh | NEMOCLAW_NON_INTERACTIVE=1 NEMOCLAW_ACCEPT_THIRD_PARTY_SOFTWARE=1 bash\n```\n\nIf you use nvm or fnm to manage Node.js, the installer may not update your current shell's PATH.\nIf `nemoclaw` is not found after install, run `source ~\u002F.bashrc` (or `source ~\u002F.zshrc` for zsh) or open a new terminal.\n\nWhen the install completes, a summary confirms the running environment:\n\n```text\n──────────────────────────────────────────────────\nSandbox      my-assistant (Landlock + seccomp + netns)\nModel        nvidia\u002Fnemotron-3-super-120b-a12b (NVIDIA Endpoints)\n──────────────────────────────────────────────────\nRun:         nemoclaw my-assistant connect\nStatus:      nemoclaw my-assistant status\nLogs:        nemoclaw my-assistant logs --follow\n──────────────────────────────────────────────────\n\n[INFO]  === Installation complete ===\n```\n\n### Chat with the Agent\n\nConnect to the sandbox, then chat with the agent through the TUI or the CLI.\n\n```bash\nnemoclaw my-assistant connect\n```\n\nIn the sandbox shell, open the OpenClaw terminal UI and start a chat:\n\n```bash\nopenclaw tui\n```\n\nAlternatively, send a single message and print the response:\n\n```bash\nopenclaw agent --agent main --local -m \"hello\" --session-id test\n```\n\n### Model Router (Complexity-Based Routing)\n\nNemoClaw includes an optional model router that automatically picks the most efficient model for each query. Instead of sending every request to a single large model, the router uses a lightweight encoder to predict which model in a pool can handle each query correctly, then routes to the cheapest one that meets an accuracy threshold.\n\nThe router uses the [NVIDIA LLM Router v3](https:\u002F\u002Fgithub.com\u002FNVIDIA-AI-Blueprints\u002Fllm-router\u002Ftree\u002Fv3) prefill routing engine and runs on the host as a LiteLLM proxy. The sandbox reaches it through the OpenShell gateway and continues to call `https:\u002F\u002Finference.local\u002Fv1`; do not probe `localhost:4000` or `host.openshell.internal` directly from inside the sandbox.\n\n#### Enable during onboard\n\nSelect **Model Router (complexity-based routing)** during the onboard wizard, or set `NEMOCLAW_PROVIDER=routed` for non-interactive mode:\n\n```bash\nNEMOCLAW_PROVIDER=routed nemoclaw onboard --non-interactive\n```\n\nThe onboard wizard starts the router, configures the OpenShell provider, and creates the sandbox. The router process runs on the host on port 4000.\n\n#### Configure the model pool\n\nEdit `nemoclaw-blueprint\u002Frouter\u002Fpool-config.yaml` to define which models the router can choose from:\n\n```yaml\nrouting:\n  method: prefill\n  checkpoint: llm-router\u002Fcheckpoints\u002Fprefill_router_qwen08b.pt\n  tolerance: 0.20\n  encoder: Qwen\u002FQwen3.5-0.8B\n\nmodels:\n  - name: nano\n    litellm_model: \"openai\u002Fnvidia\u002Fnvidia\u002FNemotron-3-Nano-30B-A3B\"\n    cost_per_m_input_tokens: 0.05\n    api_base: \"https:\u002F\u002Finference-api.nvidia.com\"\n\n  - name: super\n    litellm_model: \"openai\u002Fnvidia\u002Fnvidia\u002Fnemotron-3-super-v3\"\n    cost_per_m_input_tokens: 0.10\n    api_base: \"https:\u002F\u002Finference-api.nvidia.com\"\n```\n\nThe `tolerance` parameter controls the accuracy-cost tradeoff: 0.0 always picks the most accurate model, 1.0 always picks the cheapest, and 0.20 (default) allows up to 20 percentage points below the best for a cheaper model.\n\n#### Architecture\n\nThe router runs on the host, not inside the sandbox:\n\n```text\nSandbox (OpenClaw) ──> OpenShell Gateway (L7 proxy) ──> Model Router (:4000) ──> NVIDIA API\n                                                         └── PrefillRouter selects model\n```\n\nCredentials flow through the OpenShell provider system. The sandbox never sees raw API keys.\n\n### Uninstall\n\nTo remove NemoClaw and all resources created during setup, run the CLI's built-in uninstall command:\n\n```bash\nnemoclaw uninstall\n```\n\n| Flag               | Effect                                              |\n|--------------------|-----------------------------------------------------|\n| `--yes`            | Skip the confirmation prompt.                       |\n| `--keep-openshell` | Leave the `openshell` binary installed.              |\n| `--delete-models`  | Also remove NemoClaw-pulled Ollama models.           |\n\n`nemoclaw uninstall` runs the version-pinned `uninstall.sh` shipped with your installed CLI, with no network fetch at uninstall time.\n\nIf the `nemoclaw` CLI is missing or broken, fall back to the hosted script:\n\n```bash\ncurl -fsSL https:\u002F\u002Fraw.githubusercontent.com\u002FNVIDIA\u002FNemoClaw\u002Frefs\u002Fheads\u002Fmain\u002Funinstall.sh | bash\n```\n\nFor a full comparison of the two forms, see [`nemoclaw uninstall` vs. the hosted `uninstall.sh`](https:\u002F\u002Fdocs.nvidia.com\u002Fnemoclaw\u002Flatest\u002Freference\u002Fcommands.html#nemoclaw-uninstall-vs-the-hosted-uninstallsh).\n\nFor troubleshooting installation or onboarding issues, see the [Troubleshooting guide](https:\u002F\u002Fdocs.nvidia.com\u002Fnemoclaw\u002Flatest\u002Freference\u002Ftroubleshooting.html).\n\n\u003C!-- end-quickstart-guide -->\n\n## Documentation\n\nRefer to the following pages on the official documentation website for more information on NemoClaw.\n\n| Page | Description |\n|------|-------------|\n| [Overview](https:\u002F\u002Fdocs.nvidia.com\u002Fnemoclaw\u002Flatest\u002Fabout\u002Foverview.html) | What NemoClaw does and how it fits together. |\n| [How It Works](https:\u002F\u002Fdocs.nvidia.com\u002Fnemoclaw\u002Flatest\u002Fabout\u002Fhow-it-works.html) | Plugin, blueprint, sandbox lifecycle, and protection layers. |\n| [Architecture](https:\u002F\u002Fdocs.nvidia.com\u002Fnemoclaw\u002Flatest\u002Freference\u002Farchitecture.html) | Plugin structure, blueprint lifecycle, sandbox environment, and host-side state. |\n| [Prerequisites](https:\u002F\u002Fdocs.nvidia.com\u002Fnemoclaw\u002Flatest\u002Fget-started\u002Fprerequisites.html) | Hardware, software, and supported platforms, with any platform-specific pre-setup. |\n| [Inference Options](https:\u002F\u002Fdocs.nvidia.com\u002Fnemoclaw\u002Flatest\u002Finference\u002Finference-options.html) | Supported providers, validation, and routed inference configuration. |\n| [Network Policies](https:\u002F\u002Fdocs.nvidia.com\u002Fnemoclaw\u002Flatest\u002Freference\u002Fnetwork-policies.html) | Baseline rules, operator approval flow, and egress control. |\n| [Customize Network Policy](https:\u002F\u002Fdocs.nvidia.com\u002Fnemoclaw\u002Flatest\u002Fnetwork-policy\u002Fcustomize-network-policy.html) | Static and dynamic policy changes, presets. |\n| [Security Best Practices](https:\u002F\u002Fdocs.nvidia.com\u002Fnemoclaw\u002Flatest\u002Fsecurity\u002Fbest-practices.html) | Controls reference, risk framework, and posture profiles for sandbox security. |\n| [Sandbox Hardening](https:\u002F\u002Fdocs.nvidia.com\u002Fnemoclaw\u002Flatest\u002Fdeployment\u002Fsandbox-hardening.html) | Container security measures, capability drops, process limits. |\n| [CLI Commands](https:\u002F\u002Fdocs.nvidia.com\u002Fnemoclaw\u002Flatest\u002Freference\u002Fcommands.html) | Full NemoClaw CLI command reference. |\n| [Troubleshooting](https:\u002F\u002Fdocs.nvidia.com\u002Fnemoclaw\u002Flatest\u002Freference\u002Ftroubleshooting.html) | Common issues and resolution steps. |\n\n## Project Structure\n\nThe following directories make up the NemoClaw repository.\n\n```text\nNemoClaw\u002F\n├── bin\u002F              # CLI entry point and library modules (CJS)\n├── nemoclaw\u002F         # TypeScript plugin (Commander CLI extension)\n│   └── src\u002F\n│       ├── blueprint\u002F    # Runner, snapshot, SSRF validation, state\n│       ├── commands\u002F     # Slash commands, migration state\n│       └── onboard\u002F      # Onboarding config\n├── nemoclaw-blueprint\u002F   # Blueprint YAML and network policies\n│   └── router\u002F\n│       ├── pool-config.yaml  # Model pool and routing config\n│       └── llm-router\u002F      # LLM Router v3 submodule (prefill routing engine)\n├── scripts\u002F          # Install helpers, setup, automation\n├── test\u002F             # Integration and E2E tests\n└── docs\u002F             # User-facing docs (Sphinx\u002FMyST)\n```\n\n## Community\n\nJoin the NemoClaw community to ask questions, share feedback, and report issues.\n\n- [Discord](https:\u002F\u002Fdiscord.gg\u002FXFpfPv9Uvx)\n- [GitHub Discussions](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FNemoClaw\u002Fdiscussions)\n- [GitHub Issues](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FNemoClaw\u002Fissues)\n\n## Contributing\n\nWe welcome contributions. See [CONTRIBUTING.md](CONTRIBUTING.md) for development setup, coding standards, and the PR process.\n\n## Security\n\nNVIDIA takes security seriously.\nIf you discover a vulnerability in NemoClaw, **DO NOT open a public issue.**\nUse one of the private reporting channels described in [SECURITY.md](SECURITY.md):\n\n- Submit a report through the [NVIDIA Vulnerability Disclosure Program](https:\u002F\u002Fwww.nvidia.com\u002Fen-us\u002Fsecurity\u002Freport-vulnerability\u002F).\n- Send an email to [psirt@nvidia.com](mailto:psirt@nvidia.com) encrypted with the [NVIDIA PGP key](https:\u002F\u002Fwww.nvidia.com\u002Fen-us\u002Fsecurity\u002Fpgp-key).\n- Use [GitHub's private vulnerability reporting](https:\u002F\u002Fdocs.github.com\u002Fen\u002Fcode-security\u002Fhow-tos\u002Freport-and-fix-vulnerabilities\u002Fconfigure-vulnerability-reporting\u002Fconfiguring-private-vulnerability-reporting-for-a-repository) to submit a report directly on this repository.\n\nFor security bulletins and PSIRT policies, visit the [NVIDIA Product Security](https:\u002F\u002Fwww.nvidia.com\u002Fen-us\u002Fsecurity\u002F) portal.\n\n## Notice and Disclaimer\n\nThis software automatically retrieves, accesses or interacts with external materials. Those retrieved materials are not distributed with this software and are governed solely by separate terms, conditions and licenses. You are solely responsible for finding, reviewing and complying with all applicable terms, conditions, and licenses, and for verifying the security, integrity and suitability of any retrieved materials for your specific use case. This software is provided \"AS IS\", without warranty of any kind. The author makes no representations or warranties regarding any retrieved materials, and assumes no liability for any losses, damages, liabilities or legal consequences from your use or inability to use this software or any retrieved materials. Use this software and the retrieved materials at your own risk.\n\n## License\n\nApache 2.0. See [LICENSE](LICENSE).\n","NVIDIA NemoClaw 是一个开源参考堆栈，旨在更安全地运行 OpenClaw 始终在线的助手。它集成了 NVIDIA OpenShell 运行时，这是 NVIDIA Agent Toolkit 的一部分，提供了额外的安全性来运行自主代理。核心功能包括引导式入门、加固蓝图、状态管理、OpenShell 管理的消息通道、路由推理和分层保护。该项目特别适合需要在受控环境中部署和管理智能助手的企业或开发者使用。当前处于 Alpha 阶段，适用于早期实验和反馈收集。",2,"2026-06-11 02:55:55","top_language"]