Inoyu Apache Unomi MCP Server

The Inoyu Apache Unomi MCP Server bridges Claude and the powerful customer‑profile engine of Apache Unomi, giving AI assistants a persistent, context‑aware view of each user. By exposing Unomi’s profile store through the Model Context Protocol, developers can let Claude remember preferences, demographic data, or engagement history across sessions without writing custom persistence logic. This is especially valuable in conversational commerce, support, or recommendation scenarios where the assistant must act on up‑to‑date user information.

At its core, the server offers email‑based profile lookup and automatic creation. When a user first interacts with Claude, the assistant can query Unomi for an existing profile or create one on demand. The same mechanism powers profile property management, allowing the assistant to read and update attributes such as name, age, or subscription status. All exchanges are JSON‑encoded, keeping the interface lightweight and language‑agnostic.

The server also handles session management by generating a date‑based session ID for each interaction. This ensures that profile updates are correctly scoped and that analytics or event tracking can later be correlated with specific conversational sessions. A default scope, , is automatically provisioned for all operations, but developers can create additional scopes (e.g., per‑application or per‑region) to isolate data and control access.

Key tools exposed via MCP include:

• – retrieves the current user’s profile, optionally including segments and scores.
• – updates arbitrary properties on the user’s profile.
• and – fetch or search profiles by ID or query string.
• – defines new scopes for fine‑grained data isolation.

These tools enable a variety of real‑world use cases. In a retail chatbot, the assistant can personalize offers by reading a user’s purchase history stored in Unomi. In a support context, the bot can pre‑populate knowledge base searches with known customer details, reducing friction. For analytics teams, the session IDs and scopes allow downstream pipelines to attribute user actions back to specific conversational flows.

Because the server is built on MCP, it integrates seamlessly into any Claude workflow that supports external tools. Developers simply add the server to their configuration, and the assistant automatically discovers the available tools. The result is a declarative, extensible bridge that keeps user context consistent across devices and conversations, freeing developers to focus on higher‑level business logic rather than data plumbing.