The meeting introduced and demonstrated Laravel Boost, a new package designed to enhance AI coding agents' capabilities by giving them direct access to a Laravel application's environment, tools, and structure.
Key topics covered included installation, configuration, exploration of new files and guidelines, integration with popular editors and agents, practical use cases, debugging, and feedback workflows.
Several hands-on problem-solving scenarios were showcased, including querying databases, resolving N+1 issues, and handling application errors—all powered by Boost tooling.
The session concluded with encouragement to provide feedback to the package creator and an outlook on continued development.
Action Items
(No explicit due date – All Attendees): Try out Laravel Boost in a local Laravel application and explore available tools and integrations.
(No explicit due date – All Attendees): Provide feedback on Boost using the built-in feedback tool for the package creator, Ashley.
Installation and Configuration of Laravel Boost
Laravel Boost is added as a dev dependency using Composer.
The boost install laval command sets up the package, prompting the user to select their editors (Cloud Code, Cursor, PHPStorm, Juni) and agent (Cloud Code was used in the demo).
The MCP server is configured for each supported editor, enabling a suite of tools for AI agents.
Boost automatically detects Laravel version, adding relevant guidelines and core rules to the project.
Custom project guidelines (from the AI guidelines folder) are incorporated automatically.
New or modified files include editor-specific guideline summaries and MCP server config files.
Exploring Boost-Enabled Tools and Integrations
The installed MCP server provides tools such as database schema exploration, error logs, config key listings, running Artisan commands, querying the database, and accessing up-to-date documentation.
A Tinker tool allows agents to execute PHP code directly for tasks such as database seeding or running code samples without leaving trace files.
Integration works consistently across Cloud Code, Cursor, and Juni, although UI details may differ.
Practical Applications and Problem Solving with Boost
Demonstrated querying the database for feature items using an AI agent, which accessed the database schema and performed SQL queries via Boost tools.
Showcased the process of having the agent add new features directly to the database using the Tinker integration.
Ran a database seeder via AI agent commands, populating the application with multiple example features.
The debug bar was inspected to identify an N+1 query issue (42 duplicate queries observed).
The AI agent used route and controller analysis plus latest documentation (via search docs) to diagnose and fix the issue by implementing Laravel 12's automatic eager loading and prevent lazy loading in non-production environments.
Result after fix: database queries on the features page dropped from 40+ to 4.
Debugging and Error Handling with Boost
Removal of the application key from the environment file triggered a Laravel error, which the AI agent detected by reading server and browser logs made available via Boost.
The agent also discovered and fixed a JavaScript error affecting the code copy feature, validating its fix by re-testing the page.
Feedback and Future Direction
The feedback tool was demonstrated, sending praise and suggestions directly to the package creator, Ashley.
Attendees encouraged to submit their own feedback through their coding agent.
Plans for future expansion of Boost were mentioned, with an invitation to monitor updates and contribute feedback.
Decisions
Deploy Laravel Boost in local Laravel development environments using provided guidelines and integrations — Rationale: Enhances AI agents' capabilities and improves developer productivity by allowing direct access to application context and tools.