# Security at SixWays

> Where your data goes with each SixWays component, how releases, licenses, and images are signed, and how to report a vulnerability.

Source: https://sixways.ai/security/

## Security is the product.

Here is where your data goes, how we sign what we ship, and how to tell us when we got something wrong.

[Report a vulnerability](mailto:security@sixways.ai)

### Where your data goes.

Each component runs in a defined place. You choose which ones you use.

| Component | Runs in | What it handles |
| --- | --- | --- |
| Endpoint (open source) | Your machine | Sessions and events stay in a local database. On its own, the endpoint contacts SixWays only to check for signed updates. |
| Server | A dedicated SixWays tenant, or your infrastructure | Receives the telemetry your policy asks endpoints to send. Secrets are encrypted at rest. |
| Firewall | Your network | Prompt and response content never transits SixWays infrastructure. |
| Cloud Sandbox | Your cloud account | Sandboxes and browser terminal traffic stay in your tenant; SixWays sends signed lifecycle commands. |
| SixWays Cloud | SixWays | Licenses, billing, and your account. It does not handle agent traffic, code, or telemetry. |

### Signed, end to end.

#### Releases and updates

Release artifacts and the update feed are signed with an Ed25519 key held in a hardware security module. The endpoint verifies every update against the public key built into it, and a bad release can be recalled.

#### Licenses

Licenses are Ed25519-signed tokens that endpoints and servers verify offline. A license change can never turn off isolation.

#### Images

Sandbox and deployment images are pinned by digest, scanned before publishing, and signed, and can be verified offline against the SixWays release key.

#### Commands and evidence

Lifecycle commands are signed, approvals are bound to the exact payload, and the evidence ledger is append-only with an offline verifier that detects tampering.

### Report a vulnerability.

#### Tell us privately

Email [security@sixways.ai](mailto:security@sixways.ai) with the affected component and version, steps to reproduce, and the impact you expect. Please give us a chance to fix the issue before you disclose it publicly. Our contact details are also in [security.txt](https://sixways.ai/.well-known/security.txt).

#### Compliance

A SOC 2 Type I audit is planned for about six months after general availability, and Type II about six months later. Until then we share architecture documentation, data-flow diagrams, and security questionnaire responses on request.

### Get early access.

Tell us about your agents and your team, and we will set you up.

[Get early access](mailto:beta@sixways.ai?subject=Early%20access) [Talk to us](https://sixways.ai/contact/)