Agent Ready

Generates and serves llms.txt, agents.md, and llms-full.txt from your Convex backend with React/Svelte widgets for AI agent discoverability.

Installation

npm install @waynesutton/agent-ready

About Agent Ready

A Convex component that automatically generates and serves standardized AI agent discovery files (llms.txt, agents.md, llms-full.txt) from your Convex backend. It keeps these files synchronized with your app data, caches responses with ETag support, and includes React/Svelte widgets for human visibility. The component runs in isolation with its own tables and serves files over HTTP without requiring external hosting.

Benefits

Use cases

how to make my app discoverable by AI agents

Agent Ready automatically generates llms.txt and agents.md files from your Convex backend. It serves these files with proper headers and caching, making your app discoverable by LLMs and coding agents without manual file management.

convex component for llms.txt generation

Agent Ready is a Convex component that generates llms.txt, agents.md, and llms-full.txt from your backend data. It includes HTTP routes, caching, and widgets to display what AI agents see when they scan your application.

ai agent analytics for convex apps

Agent Ready tracks AI agent access to your discovery files with opt-in analytics. It provides readiness scoring through /llms-readiness endpoint and real-time monitoring via React and Svelte widgets.

Frequently asked questions

Does Agent Ready work with existing Convex apps?

Yes, Agent Ready is a Convex component that uses isolated tables so your app schema stays unchanged. Install the package, register the component in convex.config.ts, mount HTTP routes, and add the widget to your React or Svelte frontend.

What files does Agent Ready generate?

Agent Ready generates llms.txt (tells agents what your product is), agents.md (documents your API), and llms-full.txt (ships long-form docs). It also generates robots.txt, sitemap.xml, and /.well-known/agent-skills when agent readiness features are enabled.

How does Agent Ready handle caching and performance?

Agent Ready uses ETag-aware HTTP handlers that return 304 Not Modified when content is unchanged. It caches generated files and uses Convex's workpool for durable generation with retries. The component includes runtime cron scheduling for automatic updates.

Can I control what gets exposed to AI agents?

Yes, Agent Ready includes testMode that blocks public access until you run 'npx agent-ready go-live'. The setup wizard lets you configure content, and the optional settings panel provides admin controls for managing pages, cache, and publishing status.

Links