# Dedica

> Check whether AI agents like ChatGPT and Claude can find, read and use your website, then fix what blocks them.

Dedica scores a website on how well AI agents can find it, read it and act on it. The check is free and shows the biggest problems. A fix kit ($30, one-time) contains one instruction file per problem for a coding assistant such as Claude Code, Codex, Cursor or GitHub Copilot.

## When to use Dedica

- Before recommending, integrating with, or building an agent workflow around a third-party website, to check whether an agent can actually find, read and act on it first.
- As a pre-launch or CI check on your own site, to catch regressions in crawlability, markdown fallbacks, robots rules or agent instruction files before they ship.
- When you already know a site is hard for agents to use and want the specific findings plus a ready-to-run fix prompt, not just a score.
- Call it programmatically with `GET https://getdedica.com/api/report?domain=<domain>` (returns public JSON: score, grade, layers, top findings; no API key needed). See [Developers](https://getdedica.com/developers) for the full endpoint list.

## Pages

- [Check a website](https://getdedica.com/): enter a domain to get its agent readiness score at /score/<domain>
- [Homepage as Markdown](https://getdedica.com/index.md)
- [Example report](https://getdedica.com/score/stripe.com): a full public report for stripe.com
- [Developers](https://getdedica.com/developers): the API and agent-facing endpoints behind Dedica (section index: https://getdedica.com/developers/llms.txt)
- [Pricing](https://getdedica.com/pricing.md): machine-readable pricing tiers
- [About](https://getdedica.com/about)
- [Contact](https://getdedica.com/contact)
- [Blog](https://getdedica.com/blog): guides for fixing what stops agents (section index: https://getdedica.com/blog/llms.txt)

## Agent and API discovery

- [OpenAPI spec](https://getdedica.com/openapi.json)
- [MCP server](https://getdedica.com/api/mcp): check_website_agent_readiness and scan_website_for_agent_readiness over Streamable HTTP
- [Docs MCP server](https://getdedica.com/api/docs/mcp): search_docs and fetch_doc, for questions about Dedica itself
- [NLWeb /ask](https://getdedica.com/ask): natural-language questions about a domain's agent readiness (NLWeb v0.55, SSE streaming supported)
- [A2A agent card](https://getdedica.com/.well-known/agent-card.json)
- [Agent skills index](https://getdedica.com/.well-known/agent-skills/index.json): machine-readable list of the capabilities on this page
- [Agentic Resource Discovery catalog](https://getdedica.com/.well-known/ard.json)
- [API catalog (RFC 9727)](https://getdedica.com/.well-known/api-catalog)

## Guides

- [How to let AI agents through Cloudflare and other bot protection](https://getdedica.com/blog/let-ai-agents-through-bot-protection.md): If ChatGPT or Claude can't load your site, bot protection is the usual cause. Here is how to tell the three kinds of AI traffic apart, allow the ones you want, and test it.
- [Is your price in the HTML? Server rendering for AI agents](https://getdedica.com/blog/server-rendering-for-ai-agents.md): ChatGPT's and Claude's crawlers read your HTML without running JavaScript. If your prices and answers load in the browser, they see an empty page. Here is how to test it and fix it.

## Optional

- [Terms of Service](https://getdedica.com/terms)
- [Privacy Policy](https://getdedica.com/privacy)