Blog

There is a moment in every language's evolution when a convenience syntax grows up and becomes something worth thinking about. For BoxLang, that moment arrived with ranges. We had always wanted to be able to do this in our CFML apps, and we finally can

BoxLang 1.14 ships with one of the most developer-friendly OOP features we've built yet: local template classes. If you've ever created a throwaway .bx file just to hold a five-line helper class, this one's for you.

Every application eventually has to deal with deeply nested data. JSON API responses with payloads six levels deep. Configuration files where the key you need is buried inside an array of objects, one of which has a null for the field you thought was required. Module metadata structures that nobody wrote a schema for. Runtime introspection data shaped like a tree that grew without a plan.

BoxLang has always embraced a simple truth: the way you organize code shapes the way you think about problems. For a long time, if you needed a helper class, you needed a file. One class, one .bx file, no exceptions. That's clean and predictable, but it creates real friction when a class is tightly coupled to exactly one caller and has no business existing anywhere else.

BoxLang 1.14.0 ships a lot of exciting features -- Dynamic Sets, Ranges, Inner Classes, JSONPath navigation -- but one quietly powerful addition will change the way you think about every database call in your application: Query Transformers.

BoxLang 1.14.0 ships something that JVM developers have wanted for a long time: a true first-class Set type baked directly into the language. Not a wrapper you reach for manually, not a createObject( "java", "java.util.HashSet" ) incantation you paste from a Stack Overflow answer years ago. A real BoxSet with literal syntax, operator overloads, a full functional pipeline, change listeners, JSON serialization, and deep Java interop.

BoxLang has never stood still, but 1.14.0 is something different. This is the release where the language stops filling gaps and starts defining what a modern dynamic JVM language looks like on its own terms. Sixty-five issues closed. Four innovative language features. A formatter that has grown up. And a companion module - bx-mcp - that fundamentally changes how you operate a running BoxLang application with AI.

The BoxLang image module just landed two back-to-back releases that make it significantly more capable. 1.6.0 brought CAPTCHA generation. 1.7.0 adds four new image formats, fixes a silent write bug that has been producing PNG files regardless of what extension you asked for, and adds proper alpha channel handling for formats that don't support transparency. Let's dig in. 🚀

Today we're releasing cbMCP, the official ColdBox MCP Server — a BoxLang-only module that turns your running ColdBox application into a fully-compliant Model Context Protocol (MCP) server. Plug in any MCP-capable AI client — Claude Desktop, VS Code Copilot, Cursor, Codex, Gemini CLI, OpenCode — and your AI assistant gets live, read-only introspection across the entire ColdBox platform: routing, handlers, modules, WireBox, CacheBox, LogBox, schedulers, interceptors, and async executors. 🎯

ColdBox CLI 8.11 is here, and it's one of the most significant releases we've shipped for AI-assisted development. This release wires the CLI directly into our brand new public skills directory at skills.boxlang.io, brings our AI tooling in line with industry-wide agent conventions, and introduces a wave of quality-of-life improvements that make AI integration feel less like setup and more like infrastructure.