Skip to main content
The Lexera Protocol API is built on the Model Context Protocol (MCP). It gives AI agents and enterprise systems programmatic access to human-in-the-loop legal verification. The API is fully asynchronous and stateful, using a Server-Sent Events (SSE) architecture: you open a secure connection, dispatch a document, and retrieve certified artifacts once the human verification loop completes. Pricing is per-jurisdiction and denominated in AED. Lexera automatically resolves regional aliases (for example, UAE Federal and DIFC both map to UAE) and places a temporary hold on your corporate wallet when you route a document to a specialist.

Quickstart

Run your first statutory audit in minutes.

Authentication

Bearer tokens, key management, and HTTPS requirements.

Execution Lifecycle

SSE streams, session IDs, and JSON-RPC dispatch.

Tools Overview

All MCP tools, invocation patterns, and auth requirements.

​
How it works

1

Open an SSE stream

Connect to https://lexera.dev/mcp-server/sse with your Bearer token. The server returns a persistent stream and an endpoint event containing your session ID.
2

Dispatch a document

POST a JSON-RPC 2.0 payload to your session URL calling lexera_route_to_specialist. The server returns HTTP 202 Accepted and places a temporary AED hold.
3

Retrieve certified artifacts

Poll lexera_get_order_status over the same SSE stream. When current_status becomes verified_executed, download the law firm-certified .docx redline and report.

​
Explore

MCP Architecture

Why Lexera uses MCP, JSON-RPC 2.0 over SSE, and async stateful design.

Wallet & Pricing

AED wallet holds, per-jurisdiction tiers, and alias resolution.

Tools Overview

Reference for every MCP tool and its JSON-RPC invocation pattern.

Errors

Error signatures, streamed payloads, and how to resolve them.