Tidewave Rails

You can add Tidewave Rails to your app by running:

or by manually adding thetidewavegem to the development group in your Gemfile:

gem "tidewave", group: :development

2. Add the Tidewave MCP to your agent/editor

Add the Tidewave MCP server to your editor or MCP client configuration as the type "http" (streamable), pointing to the/tidewave/mcppath and port your web application is running at. For example,http://localhost:3000/tidewave/mcp.

As with any other MCP server, your agent will call the tools exposed by the Tidewave MCP whenever it sees fit. But you can also prompt it to call them explicitly.

Evaluates Ruby code in the context of your running application, with access to its runtime, loaded dependencies, and in-memory data, returning the result plus anything printed to standard output. It's like a Rails console for the agent.

Your agent can use it when it would rather run code than assume behavior, grounding its next step in what the running app actually does. For example, calling a method to see what comes back or reproducing a failing code path against live app state to debug it.

Runs a SQL query against your app's development database and returns the rows to the agent.

Your agent can use it to run any SQL against your development database. For example, ask it to insert some test records to see how a page looks with realistic data. Or, after a create action, the agent can verify whether the record was saved with the expected values.

Looks up the documentation for a class, method, or constant, reading from the exact gem versions locked in your app's Gemfile.lock.

Your agent can use it when it's unsure how a class or method works, so the code it generates is grounded in the docs for the exact versions of the gems your app uses, rather than training data that may be stale or a generic docs lookup that can't guarantee it matches the version your app depends on.

Returns output from your running server's log.

Your agent can use it to see what happened after a request. For example, reading the request log and backtrace when something misbehaves, or checking the log after an action to confirm the request came in with the expected params.

Returns the file and line where a class, module, or method is defined, across both your app and its dependencies.

Your agent can use it to jump straight to where a class or method is defined, by file and line, instead of grepping for it, including when the definition lives in a gem dependency.

Also, because it resolves the location from your running app instead of parsing source text, it handles metaprogramming, where a method is generated at runtime and does not appear as a literaldeffor grep to find.

Why no tools for routes, associations, etc?

Tidewave does not include tools for listing your routes, associations, etc. because agents are better off reading their respective source files, which gives agents more context and enables them to perform any necessary edit without additional tool calls.

Instead, Tidewave aims to fill in missing gaps, such as evaluating code inside your Rails app (without starting new instances) and finding source location, which can be tricky, even with grepping, due to metaprogramming and the different places Bundler can install your dependencies.

This may happen if you are compressing your responses (gzip, brotli, etc) after the Tidewave middleware runs. Usebin/rails middlewareand make sure Tidewave comes afterRack::Deflateror similar. Also look into your browser and terminal logs for any errors.

If you are using multiple hosts/subdomains during development, you must use*.localhost, as such domains are considered secure by browsers. Additionally, add the following toconfig/initializers/development.rb:

And make sure you are usingrack-sessionversion2.1.0or later.

The above will allow your application to run embedded within Tidewave across multiple subdomains, as long as it is using a secure context (such asadmin.localhost,www.foobar.localhost, etc).

If you have enabled Content-Security-Policy, Tidewave will automatically enable "unsafe-eval" underscript-srcin order for contextual browser testing to work correctly. It also disables theframe-ancestorsdirective. This is done only in the environments that Tidewave is loadead (development by default).

Tidewave is a powerful tool that can help you develop your web application faster and more efficiently. However, it is important to note that Tidewave is not meant to be used in a production environment.

Tidewave will raise an error if it is used in any environment where code reloading is disabled (which typically includes production).

You may configuretidewaveusing the following syntax:

config.tidewave.team = { id: "my-company" }

-

logger_middleware- The logger middleware Tidewave should wrap to silence its own logs

preferred_orm- which ORM to use, either:active_record(default) or:sequel

team- set your Tidewave Team configuration, such asconfig.tidewave.team = { id: "my-company" }

toolbar- controls whether the Tidewave toolbar is injected into HTML pages. Defaults totrue

A thank you to Yorick Jacquin for the initial version of this project.

bundle exec ruby -Itest test/all_test.rb

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

This is a web browser that enables your coding agent, such as Claude Code, to visit websites on your behalf and assist you in identifying bugs or creating UI test cases.

Create crafted UI components inspired by the best 21st.dev design engineers.

Bring agent evaluations, observability, and synthetic test set generation directly into your IDE for free with Galileo's new MCP server

An MCP server to help AI assistants to answer questions and generate AccelByte Extend SDK code more effectively .

MCP server for AI Diagram Maker — generate beautiful software engineering diagrams directly inside Cursor, Claude Desktop, Claude Code, or any MCP-compatible AI agent

ALAPI MCP Tools,Call hundreds of API interfaces via MCP

AI-powered SVG animation generator that transforms static files into animated SVG components using the Allyson platform

MCP server that gives AI assistants on-demand access to 1,500+ amCharts docs, ~300 code examples, and 1000+ class API references.

APIMatic MCP Server is used to validate OpenAPI specifications using APIMatic. The server processes OpenAPI files and returns validation summaries by leveraging APIMatic’s API.

One shared context layer for AI agents and humans — live API specs, DB schemas, and versioned contracts across repos so every agent and teammate works from the same source of truth.