Tag: AI

The Model Context Protocol (MCP) gives AI agents a structured way to call tools, read resources, and take actions in external systems. If you have a Laravel app, you can build an MCP server that lets agents interact with your application (creating content, querying data, triggering workflows) through a well-defined interface you control. This isn't theoretical. I've built MCP servers for production Laravel apps. Here's what I've learned.

If you've used Claude Code, Cursor, or any AI coding assistant recently, you've probably noticed they're getting better at doing things, not just generating text. The bridge that makes this possible is MCP: the Model Context Protocol. I've built two MCP servers that I use daily. One manages the blog you're reading right now. The other is a personal knowledge base called Acervo. Both are Laravel apps, and both let me manage content directly from Claude Code without ever opening a browser.

Laravel 13 dropped on March 17, 2026. If you read my What's New in Laravel 13 post, you already know the full changelog. This post isn't about that. This post is about the signal Laravel 13 sends: Laravel is now an AI-first framework, and it's not even subtle about it.

Every Laravel app has opinions. Naming conventions, custom casts, query scopes, validation patterns, how you structure form requests. All of it adds up to a codebase that works a certain way. Ask an AI to generate code for your app without that context and you get textbook Laravel that compiles but doesn't belong. The generated controller uses $request->all() instead of your form request classes. The migration uses string where you always use an enum cast. The query builds raw joins instead of using your custom scopes. It's technically correct and practically useless.

Claude Code is already fast. Point it at a Laravel project and it can read files, run tests, execute commands. But it's working blind. It doesn't know your routes, your database schema, or what's in your logs until it goes digging through files. That's where Laravel Boost comes in. Boost exposes your Laravel application's internals as MCP tools. Routes, models, schema, logs, tinker, all available to Claude Code without it having to parse files or guess. The difference is immediate: instead of Claude Code grep-ing through your route files, it just knows your routes.

Your AI assistant knows Laravel. It doesn't know your Laravel app. That's the gap Laravel Boost fills.

Laravel's AI SDK (the prism package from EchoLabs) gives you a clean, framework-native way to work with LLMs, embeddings, and vector stores. No raw HTTP calls, no third-party wrapper chaos. Here's how the pieces fit together.

Laravel's AI SDK (formerly Prism) gives you a single interface to work with multiple LLM providers. You write your code once, swap providers through config, and add fallbacks for when things go wrong. Here's how it works in practice and where each provider shines.

Laravel AI SDK (formerly Prism) ships with a first-class agent abstraction. An agent is an LLM-powered loop: it receives a prompt, decides which tools to call, processes the results, and repeats until it has a final answer. Unlike a single generate() call, agents maintain context across multiple steps and can take actions autonomously. Here's what that looks like in practice.

Laravel shipped its first-party AI SDK in February 2026. After months of community packages filling the gap, laravel/ai landed as the official way to work with LLMs, speech synthesis, and structured outputs, all with the ergonomics you'd expect from a Laravel package. It matured alongside Laravel 13 and is now stable. Here's what you need to know to start using it.