It’s been just over two years since the launch of Anypoint Code Builder (ACB), MuleSoft next-generation IDE. Built on VSCode, ACB has afforded MuleSoft customers a flexible modern development experience at a crucial time; We’re all still in the midst of experiencing the most fundamental shift in the development paradigm in decades, with the rise to prominence of vibe coding and agentic workflow.
While ACB is certainly the best place to develop for Mule using AI tools like MuleSoft Vibes, we’re fully committed to a broad range of innovations that make ACB the best place to build with MuleSoft. Alongside all of our AI enhancements, ACB has been undergoing monthly updates as part of our commitment towards rapid innovation for the core development workflow, across the entire application life cycle.
ACB February 2026 release
In this release, we’re focusing on the developer experience in Anypoint Code Builder by empowering you to build faster with visual custom metadata, intelligent DataWeave graphical mapping, frictionless dependency management, and unified workspaces that reduce the friction between migrating applications from Anypoint Studio into Anypoint Code Builder.
Workspace improvements
As development workflows grow more complex, consistency across tools becomes critical. This release reinforces a workspace-centric model that ensures reliability and reduces friction between Anypoint Code Builder and Anypoint Studio.
You can now import existing Anypoint Studio workspaces directly into ACB, with the system creating pointers to your existing files rather than duplicating them. This ensures that changes made in ACB are reflected correctly within the original workspace structure. Support for multi-root workspaces means core capabilities like DataWeave and API components continue to work as expected, even when multiple projects are organized within a single workspace.
We’ve also made project navigation more intuitive. Command palette actions are now context-aware, showing global actions such as importing projects only when appropriate, and surfacing project-specific actions like run and debug only when they apply to the file in focus. The result is a cleaner, more predictable development experience as projects scale.
Canvas usability enhancements
We’ve continued to refine the canvas to keep developers in flow and reduce unnecessary context switching. Metadata configuration tabs are now accessible directly from components within the canvas, allowing you to define input and output schemas without leaving your visual workspace.
Context menus and command palette options have also been streamlined. Actions are now dynamically filtered based on the current project or file, surfacing only what’s relevant and reducing cognitive load as you move through complex integrations.
The most significant change to the canvas in this release is the visual experience. We’re modernizing the canvas visualization to better reflect how developers expect development environments to feel. The updated experience is more fluid and less rigid, with improved visual hierarchy and context awareness that keeps focus on the flow itself rather than the tooling around it.
By reducing visual clutter and surfacing configuration directly where work happens, the canvas now aligns more closely with cutting-edge development experiences while preserving the clarity required for enterprise integration work.
Simplified error handling
Error handling has been streamlined to make issues faster to understand and easier to resolve. Previously, errors surfaced across multiple views and panels, requiring developers to hunt for relevant context. With this release, errors are consolidated into a more unified and consistent experience, making them easier to scan, interpret, and act on. This reduces cognitive overhead during development and shortens the path from detection to resolution.
Custom metadata
One of the biggest sources of friction in integration development is relying on generic connector metadata that doesn’t reflect the data you actually work with. With this release, we’re introducing a first-class custom metadata experience that puts developers in control.
You can now define and assign your own metadata directly within the canvas, supporting formats such as JSON, XML, CSV, flat files, fixed-width formats, and Java objects. By uploading schemas like XSD or JSON Schema, or simply providing example payloads, ACB automatically generates the corresponding structures for you.
A centralized management view allows you to view, edit, and delete metadata in one place. Any updates propagate automatically to every component that references that data, ensuring consistency across flows and reducing the risk of drift as integrations evolve.
DataWeave enhancements
At the heart of many MuleSoft integration is the Transform Message component, and in this release we’re significantly expanding what teams can do without dropping into custom scripts. These DataWeave enhancements close the longstanding gap between low-code ease and high-code complexity, making it faster and safer to build even the most sophisticated transformations.
We’ve introduced a broad set of graphical mapping improvements covering more than 58 real-world use cases. These improvements are complemented by MuleSoft Vibes, which can generate DataWeave transformations from intent to accelerate initial mapping, while keeping developers fully in control of the final logic.
Intelligent automapping can now match fields automatically based on name, either one level deep or recursively across nested structures. The graphical mapper also supports advanced patterns that previously required handwritten DataWeave, including root-to-root assignments, automatic array wrapping, and common envelope patterns used in API responses.
To help teams move quickly without unintended side effects, the mapper now includes override detection that warns you before a drag-and-drop action replaces existing logic. Type mismatches are handled automatically through built-in coercion, reducing runtime errors and manual cleanup. Together, these enhancements let developers focus on intent rather than syntax, while maintaining confidence in the resulting transformations.
External Libraries Management
Managing external dependencies has historically meant manually editing the pom.xml file, a process that’s both error-prone and time-consuming. With this release, external libraries can be added and managed directly through the UI.
You can now search Maven Central from within connector configuration settings and add dependencies with a single click. Whether you’re importing a proprietary library from a local system or a standard driver from a remote repository, the guided experience ensures dependencies are configured correctly so builds succeed.
A new visual management interface lets you view, edit, and remove libraries easily, keeping projects clean, understandable, and build-ready as they grow.
What’s next?
These enhancements are part of our broader push to make MuleSoft the most productive place to design, build, and operate integrations as development becomes increasingly AI-assisted.
There’s plenty more to come with MuleSoft Vibes, including the upcoming release of Template, Exchange Reuse, and Connector Support. To explore everything included in this release, review the full release notes for detailed coverage.
And to stay closely connected to what’s coming next in Anypoint Code Builder, join the February edition of In the Loop with ACBour community meetup where we share updates, previews, and guidance as the platform continues to evolve, as well as subscribing to Technically speaking, our newsletter on LinkedIn.