WSO2 Integration Platform Changelog

A development page at the project level brings every integration in a project together in one place instead of one component at a time.

The AI agent overview includes a chat interface for the production deployment, so you can try an agent without leaving the console.

Business users can discover and interact with AI agents from the User Portal alongside automations and reused components.

A test form lets you supply runtime arguments when triggering an automation manually, with filtering over past executions.

A dedicated portal for business users to find and run the automations, agents and reused components a team has published, separate from the developer console.

Define alert rules against your integrations, including status-code alerts, and manage them from the console.

gRPC services render as a browsable view of their methods and messages, the same way REST and GraphQL contracts already do.

An in-console assistant that answers product questions and navigates you to the right page, grounded in the product documentation.

Connect to Google Drive with OAuth instead of an API key.

Run content ingestion on a schedule, and optionally rerank retrieved chunks to improve the relevance of what reaches the model.

Browse a GraphQL integration's schema visually, with an explorer and clearer error reporting in the GraphiQL view.

Create MCP servers as their own component type, with automatic discovery and visualization of the tools they expose.

Bring an existing WSO2 Integrator project into the platform during project creation.

Set up retrieval-augmented generation — ingestion and retrieval — through a guided flow in the console.

Start from a catalogue of ready-made integrations instead of an empty project. (Feature preview.)

Automations report usage insights alongside execution history, so you can see how often and how long they run.

Create integrations from repositories hosted in Azure DevOps, alongside GitHub, GitLab and Bitbucket.

Generate an embeddable deploy button for a repository, so anyone can launch your integration into their own organization in one click.

Devant now supports generating MCP servers from Integrations as APIs, offering a complete solution to transform existing APIs into intelligent, AI-ready tools. With a centralized control plane, it streamlines the full MCP server lifecycle from creation to discovery providing a seamless experience for both API developers and AI agent builders.

Devant now supports executing Ballerina tests as part of the build process. This ensures that integrations are automatically validated during builds, helping developers maintain code quality and catch issues early.

Devant now includes the Cloud Editor, a web-based integration development environment powered by the WSO2 Integrator:BI. BI is a 100% open-source integration solution that delivers a powerful low-code Ballerina language development experience with AI-assisted capabilities. With support for any protocol and out-of-the-box connectivity to GenAI models, knowledge bases, systems, and AI agents, the Cloud Editor empowers developers to build integrations seamlessly within Devant.

Devant now makes it easy to create and manage secure VPN connections from Devant cloud data planes to any external network using Tailscale. It provides a sample image with a Tailscale proxy that can be deployed in projects with minimal setup.

Devant Tailscale VPN documentation, Guide on connecting your local HTTP service to Devant via Tailscale.