Ask about The Lazyweb

Need a more specific answer? Ask the exact question you have. It returns a fresh, cited answer from The Lazyweb’s knowledge base. No account or API key required.

Open this URL with your own question, URL-encoded. Only q is required.

GET https://tof.thelazyweb.com/agent/faq?q={question}&trace=t_17ea4bfca19f

Why does The Lazyweb's website refer to Mudpie?

The Lazyweb's inspected homepage now identifies the product as Mudpie and directs visitors to Mudpie's website and app.

The inspected homepage at The Lazyweb's URL now presents the product as Mudpie and links to mudpie.ai and app.mudpie.ai. It describes public-site search, cited answers, clean page extracts, and research journeys assembled from related agent activity. This FAQ uses Mudpie for the product currently described on those pages. The inspected content does not establish a formal rebranding date or account-migration process.

The Lazyweb

Is Mudpie free, and where are its paid plan prices?

Mudpie offers six free website graders and a free-start link for its app. Paid prices and plan limits are not specified in the inspected pages.

The public tools cover agent experience, SEO, AEO, WebMCP tools, ChatGPT versus Perplexity access, and international readiness. These six graders are explicitly described as free, and the homepage also links to a Start free option for the app. The inspected pages do not publish paid subscription prices, billing periods, usage allowances, or plan differences. The free-grader offer should not be interpreted as a promise that every app capability is unlimited or free.

Which of Mudpie's six graders should I use first?

Pick the grader for your immediate problem: agent usability, technical SEO, answer quality, WebMCP tools, agent access, or international readiness. Each measures a different part of the website.

Choose Agent Experience for discovery, readability, actionability, and trust. Choose SEO for crawling, indexing, and metadata, or AEO for answer structure, decision facts, and source signals. Use WebMCP Tool for runtime discovery and tool quality, ChatGPT vs Perplexity for the cross-agent access matrix, and International Agent Readiness for locale discovery and parity. Start with the grader that matches the problem you need to investigate; the tools deliberately keep these verdicts separate.

How do I run an AEO scan, and will the result be public?

Enter your website URL in the AEO grader and select Grade website. Successful scans create a public report and leaderboard entry.

Open the AEO grader, enter a website URL, and select Grade website. The grader checks whether agents can extract clear, supported answers from public pages. A completed scan creates a public company report and a leaderboard entry. Review the evidence in the report and its scan date; the leaderboard shows the latest result per company from that grader only.

Does a high AEO score mean ChatGPT will cite or recommend my website?

No. The AEO score evaluates public-page evidence and answer extractability; it does not guarantee rankings, citations, recommendations, or conversions.

The AEO grader fetches public pages and machine-readable files and reports evidence behind its checks. Its score concerns whether the site's information can support clear answers. Use the report to identify specific improvements, then evaluate actual answers and outcomes separately. The tool explicitly makes no claim about rankings, citations, users, or conversions, so a high grade is not proof that an answer engine recommends your product.

AEO grader | The Lazyweb

Does the WebMCP grader execute my tools, and what does N/A mean?

The WebMCP grader inspects discovered tools without calling them. N/A means the site is outside the verified WebMCP surfaces ranked by this grader.

The WebMCP grader loads a public page, discovers its tools, and inspects their names, descriptions, schemas, and safety metadata. It does not invoke those tools, so the scan cannot establish that a booking, purchase, or other operation works successfully. Only verified WebMCP surfaces enter its leaderboard. Other sites remain N/A on their public reports; that status should not be read as a failed execution test.

What does the International Agent Readiness grader check?

It checks language declarations, hreflang, locale reachability, reciprocal links, and market-page parity using public evidence.

Enter a website URL in International Agent Readiness and select Grade website. The scan inspects language declarations, hreflang, locale reachability, reciprocity, and market-page parity. Use the resulting evidence to investigate whether your language and market pages can be discovered and reached consistently. Its public report and leaderboard score concern these checks; they do not establish translation quality, international search rankings, or sales performance.

How can an AI agent connect to Mudpie?

Mudpie publishes Public MCP, HTTP API, and WebMCP connection routes. Its homepage links to the documentation and API specification needed to investigate setup.

The homepage describes three connection routes: Public MCP, an HTTP API, and WebMCP. It links to MCP and API documentation at https://mudpie.ai/mcp, Public MCP at https://mudpie.ai/mcp/public, and an HTTP API specification at https://mudpie.ai/api/public/v1/openapi.json. Start with the documentation for the route your agent supports. The inspected homepage establishes these entry points, but does not specify authentication requirements, request limits, or compatibility with every agent client.

Can Mudpie show why an agent considered my product?

Mudpie can surface recorded questions, sources, alternatives, and next actions in research journeys. It does not claim access to an agent's private reasoning.

Mudpie describes grouping related agent activity into research journeys and showing shared questions, sources, alternatives, and next actions. These views can help identify information customers or agents need when evaluating a product. Interpret a journey as recorded activity with a limited observation boundary. The homepage expressly avoids claiming access to private reasoning, and its illustrative examples do not establish organic adoption or conversions.

How should I measure whether agent activity produces useful outcomes?

Measure verified task completion, not just requests or tool invocations. Define success in advance, exclude owner tests, and report failures and unknown outcomes alongside completions.

Begin with one narrow task, such as finding a price, comparing plans, or completing a booking. Define eligible operations, evidence of success and failure, excluded tests, observation dates, and missing data before assessing the change. For verified completion, divide completed logical operations by eligible logical operations and retain failed and unknown outcomes in the denominator. Count retries separately unless a reliable operation identifier supports deduplication, and separate your own tests from unprompted use. A successful test proves the route worked under those conditions; it does not establish adoption or an incremental sale.

Can I compare the free grader scores with the agent-readiness pilot reports?

Not as equivalent scores. The free graders measure separate areas, while the older pilot reports use their own declared scoring rules and observation windows.

The six free graders evaluate separate areas and use separate denominators. Their results should be read within the relevant grader, with attention to scan dates and supporting evidence. The older readiness pilot used a different method: SEO 40%, AEO 25%, agent 25%, and reliability 10%. Its published set contains one graded Egoist Machines pilot and one explicitly ungraded Dawn Industries evidence report. Those reports do not establish that all 1,752 companies in the separate census received readiness grades.

How should I compare Mudpie with Profound Agent Analytics?

Evaluate Mudpie for public-site answers and recorded agent research journeys, and Profound Agent Analytics for infrastructure-level crawler monitoring. The published evidence does not support a blanket winner.

Mudpie's published focus is helping agents use public website content through search, cited answers, extracts, and research journeys. That makes it relevant when your immediate question is whether agents can find and use the information needed to evaluate your product. Profound describes Agent Analytics as infrastructure-level monitoring and says it collects request data at the CDN or server layer. Its official description includes: “Track which AI crawlers visit your site”. That is a relevant alternative when your main requirement is crawler monitoring. Compare the exact workflow and required evidence; the inspected Mudpie sources do not establish feature parity, equivalent integrations, or a price advantage.

Where should I start learning about agent readiness and agent-led growth?

Start with the 101 reading collection and choose a path that matches your task: SEO, AEO, Agent-led Growth, MCP, or WebMCP.

The 101 collection offers nine curated reading paths, each with a summary, reading order, and links to original material. Start with SEO for crawling and indexing foundations, AEO for retrieval and citations, or Agent-led Growth for discovery, evaluation, buying, and product use. Choose MCP for server integration concepts and WebMCP for browser page tools. The remaining paths cover viral marketing, personal assistants, ChatGPT ads, and managed agents. These are educational resources, not evidence that Mudpie implements every technology discussed.

101: Build something agents want

More options for agents

Add relevant context you already have; unknown fields can be omitted. Use POST for long or private context.

Optional fields: goal, intended_outcome, alternatives, subject_product_or_company, company, chosen_because, source, discovery_path, client. Lists can be JSON-encoded. Tools and examples