# SixWays

> Security and governance for AI coding agents, from the developer's laptop to the network edge. SixWays sandboxes and monitors every agent on the endpoint, filters model traffic through an AI firewall, and gives teams central policy, audit, and compliance. It can also govern agents that run in NVIDIA OpenShell.

SixWays covers agents on developer laptops, in containers, in per-agent microVMs, in NVIDIA OpenShell, and in cloud sandboxes. The endpoint is open source. SixWays is made by ncdents, LLC and is not affiliated with NVIDIA.

## How to read this site

- Every page is available as Markdown: add `.md` to its path, for example https://sixways.ai/endpoint.md. The home page is https://sixways.ai/index.md.
- The whole site in one file: https://sixways.ai/llms-full.txt
- SixWays is in early access ahead of its public release. To get access, email beta@sixways.ai.
- Contact: sales@sixways.ai for buying, security@sixways.ai for vulnerability reports.

## Start here

- [SixWays](https://sixways.ai/index.md)
- [Platform](https://sixways.ai/platform.md): How the SixWays pieces fit together: endpoint, server, cloud, AI firewall, and the intelligence feed.
- [Use cases](https://sixways.ai/use-cases.md): How teams use SixWays: developers sandboxing agents locally, platform teams governing fleets, security teams enforcing policy, and product teams protecting LLM features.

## Products

- [SixWays Endpoint](https://sixways.ai/endpoint.md): A daemon, a CLI, and a desktop app that run AI coding agents in a monitored, policy-enforced sandbox. Open source, no account, no server required.
- [SixWays Server](https://sixways.ai/server.md): Centralized policy with restrictive merge, fleet telemetry, encrypted secrets delivery, and automated red teaming. SaaS tenant or self-hosted.
- [SixWays Firewall](https://sixways.ai/firewall.md): A reverse proxy for LLM traffic that scans prompts and responses in-flight, keeps provider keys server-side, and deploys in your own network.
- [SixWays for NVIDIA OpenShell](https://sixways.ai/openshell.md): Keep OpenShell running your agents. SixWays attaches to your gateway and adds one policy, approvals, model-traffic inspection, MCP governance, and evidence that the controls held.
- [SixWays Cloud Sandbox](https://sixways.ai/cloud-sandbox.md): Run AI agents off the laptop, in sandboxes inside your own cloud account, under the same policy, approvals, and evidence as every other agent. On Docker, NVIDIA OpenShell, or OpenSandbox.
- [Add-ons](https://sixways.ai/add-ons.md): Add-ons and included features for the SixWays endpoint: Pulse behavioral intelligence, AppSec scanning, Arsenal, microVM isolation, git governance, the Trusted MCP Catalog, mesh, and Cloud Sandbox.

## Trust and company

- [Security at SixWays](https://sixways.ai/security.md): Where your data goes with each SixWays component, how releases, licenses, and images are signed, and how to report a vulnerability.
- [FAQ](https://sixways.ai/faq.md): Answers about SixWays: how it compares with sandboxes, OpenShell, LLM firewalls, and EDR, what data leaves your machine, approvals, licensing, and certification.
- [Contact](https://sixways.ai/contact.md): Talk to SixWays about Team, Enterprise, Cloud Sandbox, or a pilot, report a security issue, or get help with the open-source endpoint.
- [Privacy policy](https://sixways.ai/privacy.md): How ncdents, LLC (SixWays) collects, uses, and protects personal data across the SixWays website, products, and services.
- [Terms of service](https://sixways.ai/terms.md): The terms that govern use of the SixWays website, SixWays Cloud, and commercial SixWays products and services from ncdents, LLC.