AI Rule Engine

AI Rule Engine: an MCP server for your business logic

AI Rule Engine is a rule-driven automation platform that runs as a Model Context Protocol (MCP) server, letting AI agents and assistants securely execute your business rules as tools. Instead of copying data between systems by hand, you connect an AI client to your environment and let it trigger your automations, read approved files, and inspect execution logs — all under access controls you define.

Connect any MCP-compatible AI client

AI Rule Engine exposes a remote, HTTP-based MCP endpoint that works with the AI tools your team already uses:

• Claude, Claude Desktop, and Claude.ai — add the MCP server and sign in to approve access.
• Cursor — one-click "Add to Cursor" setup.
• Visual Studio Code — one-click "Install in VS Code".
• ChatGPT and any other client that supports remote MCP servers and custom headers.

Each rule set you allow appears to the AI as a callable tool. The assistant can run it on your behalf, read the results, and use them inside a larger task.

Secure by design: scoped tokens and OAuth 2.1

The MCP server is off until an administrator enables it for an environment. Access is then granted through scoped connection tokens — one per AI agent or integration. Each token controls exactly what a client can reach:

• Rule sets — all rule sets, or a specific allow-list.
• File access — off, all stored files, or read-only access limited to a chosen storage folder.
• Log & run access — whether execution runs and logs can be read.
• Expiration — optional expiry for automatic credential rotation.

Clients can also authorize through a guided OAuth 2.1 consent flow — no token copying required. You approve the connection, choose its scope, and revoke it at any time.

A complete rule and automation engine

Beyond MCP, AI Rule Engine lets you design rule sets from conditions and actions, manage multiple environments, and publish reusable extensions and templates from a single workspace. Build your logic once and run it from your apps, schedules, or AI agents.

Runs in your Azure subscription

Deploy AI Rule Engine as a dedicated instance in your own Azure subscription for full control over data and execution, or use shared hosting to get started quickly. Your rules, files, and logs stay within the boundaries you set.

Key capabilities

• Built-in Model Context Protocol (MCP) server for AI agents
• Rule sets exposed as AI-callable tools
• Per-token, least-privilege access control
• OAuth 2.1 guided authorization and revocation
• Read-only, folder-scoped file access for AI clients
• Execution run and log access with paging
• One-click setup for Cursor and VS Code; Claude Code and config-JSON for any client
• Reusable extensions and template marketplace
• Deploys to your own Azure subscription