tidewave-ai/tidewave_rails

What it solves

It bridges the gap between AI coding agents and a running Ruby on Rails application. Instead of agents relying solely on static code analysis or outdated training data, this tool allows them to interact with the live runtime environment to verify behavior, debug issues, and access accurate documentation.

How it works

It functions as a Model Context Protocol (MCP) server that integrates into a Rails app via a gem. It exposes a set of tools that an AI agent can call over HTTP to interact with the application's internal state, database, and logs while the app is running in development mode.

Who it’s for

Developers using AI coding agents (such as Claude Code, Cursor, or VS Code extensions) to build and maintain Ruby on Rails applications.

Highlights

• Runtime Execution: Allows agents to evaluate Ruby code in the app's context, acting as a remote Rails console.
• Database Access: Enables agents to run SQL queries to verify data or insert test records.
• Live Log Streaming: Lets agents read server logs to debug request failures and backtraces.
• Precise Documentation: Fetches documentation based on the exact gem versions locked in the project's Gemfile.lock.
• Source Location Mapping: Finds the exact file and line of classes and methods, including those generated via metaprogramming or located within dependencies.