Rails Integration
The ask-rb ecosystem offers two gems for working with Rails:
ask-rails — Add AI capabilities to your Rails app for your users. Provides generators, file conventions, and a railtie that make ask-agent and ask-graph feel native in Rails. Define agents in app/agents/, workflows in app/workflows/, compose tools from the ecosystem, and build user-facing AI features. Get started
ask-rails-harness — An admin AI copilot mounted inside your Rails app. Gives AI agents safe, controlled access to your database, code, and logs for internal development, debugging, and ops work. Ships 7 Rails-aware tools (6 from the generic ask-ruby-harness plus RouteInspector) and an admin chat UI at /ask. Set up
Both gems build on ask-agent, which provides the core agent loop. You can use one, the other, or both. The harness family also has a language edition for any Ruby project — see Ruby Harness.
ask-rails (user-facing AI)
| Page | What’s covered |
|---|---|
| Getting Started Guide | Install, define agents, add tools, stream responses |
| Actions | Operations callable from any channel — web, Slack, voice |
ask-rails-harness (admin copilot)
| Page | What’s covered |
|---|---|
| Setup & Generators | Install, configure, and generate |
| Database Tools | SchemaGraph, RouteInspector, QueryDatabase, and more |
| Persistence | ActiveRecord-backed agent sessions |
| Error Services | SolidErrors, Sentry, and Honeybadger integration |
| Agent Integration (MCP) | Connect Claude Code, Cursor, and MCP agents to your Rails app |