> ## Documentation Index
> Fetch the complete documentation index at: https://docs.sengol.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Sengol

> Signed, offline-verifiable evidence for AI agents in regulated institutions.

Sengol checks what your AI agents say against your policy book, and writes
each verdict as a signed, hash-chained record. Months later, an examiner can
verify that record on a laptop without Sengol installed.

Sengol is a self-hosted appliance. One install is one tenant, inside your own
network.

## How it works

**1. Instrument**

Add the Sengol SDK to your agent, send it OpenTelemetry spans, or call the REST API.

**2. Evaluate**

Evaluators score each response against a [ControlBook](https://docs.sengol.io/concepts/controlbook.md): your regulations, expressed as configuration.

**3. Gate**

`sengol gate` blocks a release in CI when the agent misses its thresholds. Runtime guardrails can withhold a failing response.

**4. Prove**

Export an evidence pack. Anyone can check it with the open-source `sengol-verify`.

## Pick your journey

> [card] [Platform operator](https://docs.sengol.io/journeys/platform-operator.md)
>
> Install the appliance, load the licence, connect identity, assign roles.

> [card] [AI developer](https://docs.sengol.io/journeys/ai-developer.md)
>
> Evaluate your agent in code, add guardrails, and gate releases in CI.

> [card] [Compliance and risk](https://docs.sengol.io/journeys/compliance-officer.md)
>
> Choose regulations, write the ControlBook, generate reports.

> [card] [Human reviewer](https://docs.sengol.io/journeys/reviewer.md)
>
> Work the review queue and record decisions.

> [card] [Auditor or examiner](https://docs.sengol.io/journeys/auditor.md)
>
> Verify an evidence pack offline, without trusting the vendor.

> [card] [Just looking](https://docs.sengol.io/quickstart.md)
>
> Run the full loop on your laptop in about 15 minutes.

## What Sengol does not do

* It does not hold or proxy your model-provider keys.
* It does not train models. It governs a judge model that your own platform trains.
* It is not multi-tenant SaaS. A bank that needs hard separation runs a separate install.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.