Celigo platform 2026.3.1 release notes

Tool builder is a low-code, Flow-Builder-like experience for creating tools. Define tool logic visually using familiar constructs such as imports, lookups, branching, and transformations.

Each tool defines a clear contract through input and output schemas, created using JSON Schema and mapped using the mapper. Tools can be validated using mock input data before deployment, making them easy to test, reuse, and compose.

MCP server exposes APIs and integrations to AI agents with governed access, scoped security, and audit logging.

Tools can be used in different Celigo experiences depending on the integration logic.

Support for using tools directly within flows and APIs will be introduced in a future release.

The Tool step provides deterministic execution of tool logic with structured inputs and outputs. Inputs and outputs are validated against the tool’s schemas to ensure predictable behavior across integrations.

Connections can be overridden at runtime, allowing the same tool to be reused across environments without modification. A built-in Run/Preview experience enables safe iteration during configuration.

Tools are discoverable and manageable from a centralized list, with audit logging to track changes and support governance as usage scales.

Build once. Use anywhere. Run optimally for each context.

File-based imports now make generated file behavior clearer by replacing the less discoverable Skip aggregation advanced setting with explicit file-generation options. Builders can now choose whether files are generated as one file per flow run or one file per page of records, making it obvious how many files will be created and removing ambiguity around the previous default behavior. The setting will be migrated as follows:

Webhook listeners now support an optional setting, Path to records in webhook request body, that extracts the record array from large payloads before pages are created, so requests up to 50 MB can be accepted as long as the extracted records can be chunked into pages under the 5 MB page limit. After extraction (including after any parsing, such as XML to JSON), the export applies paging to the extracted records so they are split into smaller chunks that fit within page size and request limits, and an optional Include parent data setting copies data outside the records path into _PARENT on each record to preserve header-level metadata.

Webhook listeners now support HMAC-based webhook challenge flows required by applications such as Zoom, by allowing you to build a custom challenge response in the Advanced field editor (AFE) using handlebar expressions with access to the webhook key. The webhook will then generate the required signed response for endpoint validation without workarounds, while keeping key usage controlled for security.

The platform now supports much wider CSV files by increasing the maximum column limit from 500 to 2,000 in the core CSV parser and generator. This enhancement applies to all connectors and flow steps that read or write CSV (including file-based, cloud storage, HTTP-based, and bulk data scenarios), enabling enterprise datasets with high field counts to be imported and exported without hitting the previous cap.

We’ve introduced advanced concurrency management settings to the DynamoDB connector for enhanced performance tuning and rate-limit resilience. You can now configure the target concurrency level to optimize throughput, and view the current concurrency level for better visibility. A new Borrow concurrency from option allows you to share concurrency pools across connections. Additionally, we’ve added an Auto-recover rate limit errors toggle to automatically retry operations when rate-limiting issues are detected. These updates provide greater flexibility and stability when handling large volumes of data with DynamoDB.

We’ve added support for Service Principal authentication using Microsoft Entra ID (formerly Azure AD) to the Microsoft SQL connector. This enhancement provides a more secure and scalable way to connect, especially for enterprise and automated integrations.

You can now choose between two authentication types:

This update enables secure, non-interactive, app-level access to your MS SQL databases without relying on individual user credentials.

Partners can now configure a custom redirect URL for OAuth 2.0 connections, allowing merchants to be redirected back to the partner’s application after completing third-party authorization, instead of landing at the Celigo platform sign-in page.

This enhancement is designed specifically for embedded and customer-facing OAuth flows.

The OAuth framework now allows connectors to include HMAC signatures as part of the authorization URL structure. OAuth handling now support custom redirect parameters beyond the traditional redirect_uri.

The Snowflake-optimized bulk load now supports loading datasets larger than 1 GB by automatically splitting the upload into multiple files up to a maximum file size and loading them in sequence, aligning with Snowflake’s guidance to use smaller files while avoiding prior driver memory constraints. The import is unchanged aside from a new optional Load data as available checkbox (shown when Use optimized bulk load is enabled), which starts loading each file as soon as it is created instead of waiting for all files to be generated, helping reduce credits while avoiding the delay of full pre-aggregation. If you use post-aggregation scripts, the post-aggregate input is updated to support multiple files (the input now reflects a collection of file results rather than a single file).