> For the complete documentation index, see [llms.txt](https://docs.visdom.virtuslab.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.visdom.virtuslab.com/security-summary.md).

# Security summary

| Requirement                           | How it is met                                                                                                                                                                                                          |
| ------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| No outbound internet connections      | Full on-premises operation; the only model traffic goes through your LLM gateway, and models can be served entirely on-premises. Updates arrive as signed bundles into your registry                                   |
| No data passed to the vendor          | Zero outbound telemetry. The platform does not call home; code, prompts and metrics stay in your infrastructure                                                                                                        |
| Transport encryption                  | TLS 1.2+/1.3 on all interfaces; mTLS between the controller and agent pods                                                                                                                                             |
| Authentication and authorisation      | OIDC/SSO and OAuth2, MFA, granular RBAC with per-project and per-repository permissions, unique accounts, blocking rather than deletion                                                                                |
| Secrets                               | Your secret store as the source of truth, AES-256-GCM at rest, KMS/BYOK, secret substitution (the agent never sees a value), rotation without restarts                                                                 |
| Agent isolation                       | One pod per agent, NetworkPolicy default-deny, transparent egress proxy with an allowlist, no access to production                                                                                                     |
| Protection against leaking to a model | A pre-LLM gate scanning for secrets and sensitive data, stopping with an alert; secrets redacted in the audit record                                                                                                   |
| Audit                                 | A complete, signed (Ed25519), chained (SHA-256), append-only record of all agent activity and configuration changes, with actor identity and timestamp; verifiable offline; exportable to SIEM; configurable retention |
| Attribution and traceability          | Line-by-line attribution of AI-written code to model and prompt, linked to commits and pull requests, the evidence base for internal policy on generated code and for external frameworks that require traceability    |
| Penetration testing                   | The architecture is built for it: the attack surface is limited to APIs behind OIDC, and findings are addressed as part of a deployment                                                                                |


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.visdom.virtuslab.com/security-summary.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.