# amahub

> amahub is a managed workspace for AI agents: a company's people delegate real work to agents that run in an isolated cloud sandbox, reach only the company tools each person is already allowed to use, and leave an audit trail of every read, action and dollar. Built for regulated organisations.

How to read these docs:

- "Concepts" explains the nouns — tasks & sessions, processes, automations, connectors, skills.
- "Administration" is what an org admin configures — members & groups, models & limits, guardrails & DLP, alert routing, usage & spend.
- "Security" and "Self-hosting" describe the runtime (sandbox, egress proxy, data classes) and how to run amahub or its runners inside your own network.
- The product is hosted (EU region) or self-hosted; connectors are official MCP servers plus custom ones with vaulted org credentials.

## Documentation Sets

- [Abridged documentation](https://docs.amahub.io/llms-small.txt): a compact version of the documentation for amahub, with non-essential content removed
- [Complete documentation](https://docs.amahub.io/llms-full.txt): the full documentation for amahub

## Notes

- The complete documentation includes all content from the official documentation
- The content is automatically generated from the same source as the official documentation

## Optional

- [amahub website](https://amahub.io/): What amahub is, for whom, how it is controlled, pricing.
- [Trust & security](https://amahub.io/trust): Certification status, sub-processors, data handling, exit terms.