Xybern Documentation¶
Xybern is the authorisation and privacy layer for enterprise AI agents - plus vertical AI products built on top of it.
Products¶
- Authorisation Layer - every agent action (tool calls, delegations, API requests) is intercepted, evaluated against your policies, and allowed or denied before execution, with a full audit trail. Includes the MCP Server, LLM Gateway, and Framework Integrations.
- Authorised Agents - domain-specialised AI agents (Banking, Legal, Project Management, Fraud) with skills and automations, every action governed by the Authorisation Layer.
- Redact - a drop-in privacy proxy that anonymises PII before it reaches any LLM and de-anonymises the response. Change one line: the
base_url. - Scroll - an Arabic-native regulatory workflow platform for UAE and Saudi firms: read Arabic documents, prepare government submissions, gate them for human approval, and seal them in a tamper-proof vault.
Quick Navigation¶
- Getting Started - Base URL, authentication, SDK setup
- Authorisation Layer - the core enforcement API (+ MCP Server, LLM Gateway, Framework Integrations)
- Authorised Agents - Banking, Legal, Project Management, Fraud agents
- Redact - PII anonymisation proxy
- Scroll - Arabic regulatory workflows
- Resources - errors, caching, human-in-the-loop
Base URL: https://www.xybern.com/api/v1