# February 10

## Connectors

### **Updates on the MCP Server Trigger**

#### **Support for External JWT**

The MCP Server Trigger now supports authentication using **External JWT**, allowing validation of tokens from external identity providers (such as Auth0, Azure AD, and Okta) through an `externalJWT` header.

By centralizing authentication in an external IdP, you eliminate the need to manage isolated credentials and ensure a unified audit trail.

<a href="https://app.gitbook.com/s/EKM2LD3uNAckQgy1OUyZ/triggers/web-protocols/mcp-server" class="button primary">Learn more</a>

<figure><img src="https://4091659634-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FE9x0KHkAmxzJGlYPs9aE%2Fuploads%2FvCHDdfdiNhXJWevWV1WR%2F(V2)%20EN%20-%20MCP%20Server%20Trigger%20-%20External%20JWT.gif?alt=media&#x26;token=109fa45a-9587-484d-b66f-3d871e586fa3" alt=""><figcaption></figcaption></figure>

#### **Enhanced compatibility with AI agents**

We updated the **MCP Server Trigger** response to align with official MCP protocol specifications. Responses are now encapsulated inside the **`content`** field, enabling seamless, plug-and-play integration with AI agents.

This ensures accurate data interpretation and higher reliability for your AI-driven automation workflows.

### **Parquet connectors update: Improved JSON normalization**

We've updated the [**Parquet File Reader**](https://app.gitbook.com/s/EKM2LD3uNAckQgy1OUyZ/connectors/files/parquet-file-reader) and [**Stream Parquet File Reader**](https://app.gitbook.com/s/EKM2LD3uNAckQgy1OUyZ/connectors/files/stream-parquet-file-reader) connectors to improve JSON output when reading data using Parquet tools. The improvements focus on transforming raw data into readable formats with new parameters:

* `convertDateFields`: Enables `DATE`/`TIMESTAMP` fields to be converted from epoch/serial numbers to ISO-8601 strings.
* `dateFieldPaths`: Specifies fields if the schema lacks logical type declarations.
* `decodeBase64Fields`: Scans the output JSON to decode valid Base64 strings into UTF-8 text, replacing raw binary data with readable content.

## Platform Improvements

### **Experience improvements**

We’ve made several improvements to the user experience across the Build, Run, and Monitor pages, including:

* Added visual emphasis to filter dropdowns when a non-default value is selected.
* Introduced a new and improved date selector on the Monitor pages.
* Adjusted the size of input fields on the Executions and Pipeline Logs pages in Monitor.
* Updated the font weight of chart titles on the Metrics page in Monitor.
* Reordered columns in the Executions and Pipeline Logs tables.

## **Bug fixes**

* **Error on Azure Blob Storage when using Double Braces:** We fixed a bug that caused an error on the Blob Storage connector when using Double Braces expressions.
* **Error when reading Parquet files:** We fixed a bug that caused an error when reading Parquet files using Parquet File Reader or Stream Parquet File Reader.