Codename One Adds MCP Server Support, Letting AI Agents Control App UI Semantically

·1 views

Codename One, an open-source cross-platform Java/Kotlin framework, has introduced Model Context Protocol (MCP) server support via pull request #5377. The feature allows AI coding agents such as Claude Code or Codex to interact with a running Codename One simulator or desktop tool by reading a structured semantic UI tree rather than relying on screen coordinates. Developers can activate the server with a single API call, choosing either a loopback socket mode or a stdio transport for direct MCP host integration. Built-in tools let agents snapshot the UI, locate elements by label or ID, enter text, and trigger actions — all executed safely on the app's event dispatch thread. Applications can also register custom typed tools with JSON schemas, enabling agents to call app-specific logic alongside the standard UI controls.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

OpenAI: 4 Million Americans Used ChatGPT for Business Purposes in March 2026

OpenAI's report, 'Main Street Entrepreneurship, Powered by ChatGPT,' found that at least 4 million U.S. users turned to ChatGPT in March 2026 to plan, start, run, or grow a business. The figure is a conservative estimate derived from privacy-preserving classifiers applied to a randomized sample of over 300,000 consumer accounts, counting only messages clearly tied to concrete business activity. The study distinguishes between prospective entrepreneurs, who primarily use the tool for branding and product development, and active business owners, who lean on it for marketing, customer communication, and legal or compliance tasks. OpenAI frames ChatGPT as a generalist 'first hire' that helps small teams tackle diverse functions without the fixed costs of specialist staff. Active entrepreneurs were also more likely to be on higher-tier subscription plans, which the company interprets as a sign of deeper engagement as businesses grow in complexity.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Developer shares 3 bugs AI agents failed to fix while building a terminal multiplexer

A developer building 'wimux', a terminal multiplexer written largely by AI agents, documented three persistent bugs the AI could not resolve despite multiple attempts. The first bug involved corrupted terminal display after pane splits, where four AI-suggested fixes failed until the developer traced the root cause to rapid resize events overwhelming the PTY layer. The second bug caused the first keystroke after clicking a pane to be lost, which took days to diagnose and was ultimately traced to focus being stolen during sidebar clicks rather than any issue with the terminal itself. A reproduction harness confirmed the fix worked, restoring all lost keystrokes. The cases highlight that AI agents tend to apply fixes at the wrong layer of a system, and that targeted diagnostic tools — rather than repeated patches — are often what breaks a debugging deadlock.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

DEV's Big Summer Bug Smash Challenge Offers $5,000 in Prizes for Real Bug Fixes

DEV Community launched its Big Summer Bug Smash challenge on July 14, inviting developers to find and fix real bugs in existing codebases. The challenge has two tracks: 'Clear the Lineup,' which requires submitting an actual pull request with a genuine bug fix, and 'Smash Stories,' where participants share memorable debugging experiences. Over 20 winners will be selected across five prize categories, including Best Use of Sentry and Best Use of Google AI, with prizes including cash and limited-edition skateboards. Participants can work on their own projects or contribute to Forem, the open-source platform that powers DEV. Submissions are due by August 23, with winners announced on September 17.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Why Flutter's ValueNotifier Breaks Down in Complex Apps and How Signals Fix It

Flutter's built-in ValueNotifier is a simple state management tool that works well for basic use cases like toggling switches or counters, but struggles as applications grow in complexity. Because it cannot automatically observe other notifiers, developers must manually wire listener callbacks for every dependency, causing boilerplate code to grow quadratically as the number of state relationships increases. Forgetting to remove even a single listener during disposal can cause memory leaks that are difficult to trace in production. Reactive signals offer a composable alternative by automatically tracking dependencies and updating derived state without manual wiring. The BlocSignal pattern further combines signal-based reactivity with structured BLoC architecture for enterprise-scale Flutter applications.

0 comments Read more at DEV Community