# Answer Checkup

> Answer Checkup shows how a business appears in AI-generated answers for real buyer questions, plus a prioritized fix list. Answers are measured via each engine's API. Free technical scan, available now. Paid reports are coming soon: x402 (USDC) for agents, and a one-time card payment for humans.

Agents: call POST https://answercheckup.com/v1/scan for free. Coming soon: POST https://answercheckup.com/v1/report, paid by answering the HTTP 402 challenge with x402 (scheme "exact", USDC). Pricing is machine-readable at https://answercheckup.com/v1/pricing.

## API
- [OpenAPI 3.1 spec](https://answercheckup.com/openapi.json): scan, report, pricing endpoints
- [Pricing (JSON)](https://answercheckup.com/v1/pricing): $0 scan (available now); $1 for 5 prompts and $2 for 10 prompts (coming soon); human report coming soon
- [MCP server](https://answercheckup.com/mcp): streamable HTTP, tools scan_site (free) and get_report (paid, coming soon)

## Products
- Free scan: robots.txt rules for GPTBot, OAI-SearchBot, ChatGPT-User, PerplexityBot, ClaudeBot, Google-Extended and CCBot; llms.txt; schema.org JSON-LD; title and meta description; content present in raw HTML without JavaScript; sitemap. Returns a 0 to 100 score and fixes.
- Agent report (coming soon; $1 for 5 prompts, $2 for 10, over x402): 5 or 10 buyer prompts, each asked once on each enabled engine (currently: the OpenAI API with web search and the Perplexity API). Returns mention rate, citation rate, share of mentions vs up to 3 competitors, and cited URLs as JSON.
- Human report (coming soon): 25 prompts, each asked once on each enabled engine; share of mentions, cited sources and a fix list.

## Policies
- When paid reports open, payment is settled only after the report is generated. If generation fails, no payment is settled.
- Reports stay at a private, unguessable link for 30 days, then expire. Raw engine answers are cached for 24 hours.
- No ranking guarantees. API answers approximate the consumer apps but are not identical to them.
- Answer Checkup is independent and not affiliated with OpenAI, Perplexity, Google or Microsoft.

## Pages
- [Answer Checkup pricing](https://answercheckup.com/pricing): Free AI readiness scan, available now. Paid reports: coming soon. 25 buyer questions, each asked once on the OpenAI API with web search and the Perplexity API.
- [Answer Checkup for AI agents: API, MCP server and x402 payments](https://answercheckup.com/agents): Answer Checkup has an API and an MCP server. Agents pay per report over x402 in USDC: $1 for 5 questions, $2 for 10 (coming soon).
- [Answer Checkup blog](https://answercheckup.com/blog): Guides to AI SEO, generative engine optimization (GEO), and checking whether ChatGPT and Perplexity recommend your business.
- [AI SEO: how to get your business into AI answers](https://answercheckup.com/blog/ai-seo): What AI SEO is, how answer engines choose sources, and steps a small business can take, with our own site's 0/100 score as a worked example.
- [Generative engine optimization (GEO), explained](https://answercheckup.com/blog/geo): Generative engine optimization (GEO) explained: where the term came from, what the research found, what Google says, and a GEO checklist.
- [Does ChatGPT recommend my business? How to check](https://answercheckup.com/blog/does-chatgpt-recommend-my-business): How to check if ChatGPT and Perplexity mention or cite your business: a step-by-step manual test, what our own test found, and what to fix.
- [llms.txt explained: what it is and how to write one](https://answercheckup.com/blog/llms-txt): What llms.txt is, what it does and doesn't do for AI answers, and how to write one for a small business site in about 15 minutes.
- [robots.txt for AI crawlers: what to allow and block](https://answercheckup.com/blog/robots-txt-for-ai-crawlers): Which AI crawlers read your site, which are for search and which for training, and copy-paste robots.txt rules for each choice.
- [Can AI crawlers see my site? Six checks to run](https://answercheckup.com/blog/can-ai-crawlers-see-my-site): How to check if GPTBot, OAI-SearchBot and other AI crawlers can read your website: six checks, curl commands for each, and how to fix what fails.
- [Do AI crawlers render JavaScript? How to check](https://answercheckup.com/blog/javascript-ai-crawlers): Why client-side rendered pages can look empty to AI crawlers, how to test yours with curl and view-source, and the fixes: SSR, static generation, prerendering.
- [What share of voice in AI answers looks like for a well-known brand](https://answercheckup.com/blog/ai-share-of-voice): We asked AI answer engines 25 buyer questions about email marketing tools. Here is how often a well-known brand was named, who else was, and what share of voice means.
- [Which sites AI answers cite when people ask about email marketing tools](https://answercheckup.com/blog/ai-answer-cited-domains-email-marketing): We logged the sources AI answer engines cited for 25 questions about email marketing platforms. The brand's own site led, then a review site, then a competitor.
- [Entity signals: why AI can describe some brands and not others](https://answercheckup.com/blog/ai-entity-signals): Why AI answer engines describe some brands clearly and say 'I couldn't find it' about others. A side-by-side of a well-known brand at 70/100 and our own site at 25/100.
- [About Answer Checkup](https://answercheckup.com/about): Answer Checkup checks whether AI answer engines can find and mention a business. Who we are, how the free scan works, and what we do not promise.
- [Contact Answer Checkup](https://answercheckup.com/contact): How to reach Answer Checkup: email hello@answercheckup.com for questions, corrections and security reports. We reply by email.
- [Full agent docs](https://answercheckup.com/llms-full.txt): every endpoint, the x402 flow (coming soon), MCP tools and the FAQ in one file

## Contact
- Answer Checkup, operated by Kirill Stolbushkin. Email: hello@answercheckup.com

---

# Answer Checkup: full reference for agents

Base URL: https://answercheckup.com
Everything below is generated from the Worker's own code (src/discovery.ts, src/pricing.ts, src/mcp.ts) on 2026-10-15.

## Products and prices (from https://answercheckup.com/v1/pricing)
- Free AI readiness scan (id `free_scan`): free. Endpoint `POST /v1/scan`, MCP tool `scan_site`.
- Agent AI visibility report, 5 prompts (coming soon) (id `agent_report_5`): $1, 5 prompts, each asked once per engine. Endpoint `POST /v1/report` with body {"size":5}, MCP tool `get_report`. Payment: x402 v2, scheme exact, USDC on eip155:84532 (Base Sepolia testnet), amount 1000000 (6 decimals).
- Agent AI visibility report, 10 prompts (coming soon) (id `agent_report_10`): $2, 10 prompts, each asked once per engine. Endpoint `POST /v1/report` with body {"size":10}, MCP tool `get_report`. Payment: x402 v2, scheme exact, USDC on eip155:84532 (Base Sepolia testnet), amount 2000000 (6 decimals).
- Full AI visibility report (coming soon) (id `human_report_25`): no public price yet, 25 prompts, each asked once per engine. Payment: card (Stripe Payment Link, checkout not open yet).
- Engines: OpenAI API with web search (Approximates ChatGPT search; not the ChatGPT app.); Perplexity API (Perplexity Agent API, fast preset.).
- x402 payments are settled only after the report is generated successfully. If generation fails, the payment is not settled and an error is returned.
- Reports are kept 30 days. Engine answers are cached 24 hours.
- Answers are measured via each engine's API. They approximate, but are not identical to, what users see in the consumer apps. No ranking guarantees.

## Free scan
`POST https://answercheckup.com/v1/scan` with JSON `{"url":"https://example.com"}` (also `GET https://answercheckup.com/v1/scan?url=https://example.com`).
Checks: robots.txt rules for 7 AI crawlers; llms.txt presence and validity; schema.org JSON-LD types; title and meta description; main content in raw HTML (no JavaScript); sitemap; score and prioritized fixes.
Returns JSON with `score` (0 to 100), `grade`, `checks`, `scoreBreakdown` and `fixes` (each with priority high, medium or low).
Rate limits: 20 scans per network per hour, 10 per scanned domain per hour (HTTP 429 with retry-after).

## Paid agent report (x402, coming soon)
Paid reports are not open yet. This is how the paid flow works once they open.
1. `POST https://answercheckup.com/v1/report` with JSON `{"url":"https://example.com","brand":"Example Co","size":5}`. Optional: `prompts` (up to `size` strings), `category`, `competitors` (up to 3).
2. Invalid input returns 400 before any payment is asked for.
3. Without payment the answer is HTTP 402 with a base64 JSON `PAYMENT-REQUIRED` header (x402 v2). `accepts[0]` has `scheme` "exact", `network` "eip155:84532", `amount` "1000000" for size 5 or "2000000" for size 10, `asset` 0x036CbD53842c5426634e7929541eC2318f3dCF7e (USDC) and `payTo`.
4. Sign the payment (for example with @x402/fetch or @x402/evm) and repeat the same request with the `PAYMENT-SIGNATURE` header.
5. HTTP 200 returns `{"reportUrl", "expiresAt", "report"}` and a `PAYMENT-RESPONSE` header. HTTP 502 means generation failed and the payment was not settled. A payment can be used once (409 on reuse).
x402 payments are not open yet; the endpoint currently runs on the Base Sepolia test network only.

## MCP server
Streamable HTTP at https://answercheckup.com/mcp (stateless). Server card: https://answercheckup.com/mcp/server-card. Tools:
- `scan_site` {url}: the free scan above.
- `get_report` (paid, coming soon) {url, brand, size: 5|10, prompts?, category?, competitors?, payment?}: call without `payment` to get the x402 requirements, then again with `payment` set to the base64 PAYMENT-SIGNATURE value.

## Other endpoints
- `GET https://answercheckup.com/v1/pricing`: pricing JSON.
- `GET https://answercheckup.com/v1/reports/{id}` (JSON) and `GET https://answercheckup.com/r/{id}` (HTML): a stored report by its private id. Not indexed.
- `GET https://answercheckup.com/openapi.json`: OpenAPI 3.1.0.

## Questions
### Is this the same as what I see in the AI apps themselves?
Close, not identical. We measure through the engines' APIs, which approximate but don't match the consumer apps. Answers also vary run to run, so paid reports (coming soon) ask many buyer questions, each once per engine, and show how often you appear across them.

### Do you guarantee I'll show up?
No. Nobody can promise that. We show where you stand and what's likely blocking you.

### What questions do you ask?
We generate buyer questions from your site and category, and you can edit them before the run.

### Do you store my data?
Only what's needed to build your report. Your report stays at a private link for 30 days, then we delete it. Raw engine answers are cached for 24 hours so repeat checks are fast. We never sell or share your data.