# May 12 ## **Agents** ### **Dynamic LLM model parameters in the Agent Component** The Agent Component now loads models and parameters dynamically. Previously, the list of available models and their parameters were fixed. Now, both are fetched directly from the platform, so the options you see in the component always reflect the current state of the catalog, including which parameters each model actually supports. The catalog refreshes every 24 hours. When new models are added to the platform, they become available in the Agent Component automatically, with no manual update required. Learn about the Agent Component ## **Platform Improvements** ### **Version Diff Viewer for Capsules** You can now compare changes across **different published versions of a Capsule**, with a clear view of what was added, removed, or modified. This makes it easier to audit versioning history and understand the impact of each Capsule update. Learn about Capsules versioning

*** ## **Connectors & Triggers** ### **Connector mocking for JSON Generator** You can now enable mocks for the **JSON Generator connector**, allowing you to create multiple example payloads with hardcoded values and simulate different input scenarios. This makes it easier to test conditional logic and validate pipeline behavior without requiring live data. The feature is available in pipelines, capsules, and Test Cases, and follows the same mocking experience as other connectors. Read more about the JSON Generator connector ### **Metadata support for File Writer and AWS S3 Storage** You can now attach custom metadata to files written with the **File Writer** connector and to objects uploaded through **AWS S3 Storage**. The File Writer returns the metadata in the step response and can optionally write a companion sidecar file alongside the output, making it easy to pass context to downstream steps. When uploading to S3, you can provide the metadata inline or point the connector to the sidecar file generated by the File Writer, and it will be sent as user metadata on the S3 object. Access the documentation: * [**File Writer**](/documentation/connectors-and-triggers/connectors/files/file-writer.md) * [**AWS S3 Storage**](/documentation/connectors-and-triggers/connectors/aws/s3-storage.md) ### **OpenAPI and cURL import in REST V2** You can now import **OpenAPI specs** and **cURL commands** directly into the **REST V2 connector**, automatically populating configuration fields such as URL, HTTP method, headers, query parameters, and request body. This reduces manual setup time and configuration errors when integrating with documented APIs. Learn about REST V2

### **Update on FUZZYMATCH function: Array support for candidates** The FUZZYMATCH function documentation has been updated to reflect that each candidate argument after the threshold can now be a scalar string or a JSON array, with elements evaluated as independent candidates. Scalars and arrays can be mixed in the same call. Read the documentation *** ## **Bug Fixes** * **Diff Viewer:** In addition to indicating that a step was added, the Diff Viewer now also displays the step's content, allowing you to see the details of the addition directly in the version comparison. * **Dynamics 365 Connector:** Fixed an out-of-memory error that occurred when loading the entities list. * **Capsules:** Fixed incorrect actions displayed for archived capsules when accessed through the sidesheet: the Restore option is now shown correctly instead of Publish and Archive. * **DynamoDB Connector:** Fixed an issue where UpdateItem, Query, and Scan operations returned a 400 error from AWS when the same connector instance processed multiple messages in sequence. * **Stream DB V3 and DB V2 connectors:** Fixed an issue where form fields were displayed out of order and misaligned when configuring these connectors; fields now render correctly and in the expected order. * **AI Assistant — JOLT Generator:** Fixed an issue where JOLT Generator instructions were displayed in Portuguese for users operating the platform in English; the content now appears in the correct language. --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://docs.digibee.com/documentation/release-notes/may-2026/may-12.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.