Loading ToolBench
MCP server for AI visibility and AEO (Answer Engine Optimization) analysis — connect via Claude or ChatGPT to analyze how your site ranks on AI platforms
Five tools with clear naming (verb_noun pattern) and detailed descriptions (150-250 chars each). All tools have input schemas with typed parameters and descriptions. However, output schemas are not documented, responses are returned as plain text strings rather than structured objects. Parameter descriptions lack constraint details (ranges, enums, formats). No error handling guidance or recovery paths documented. Missing idempotency and permission declarations.
Crawl a website and score it for AEO (Answer Engine Optimization) — how well it's structured for AI assistants to discover and cite it. Returns a 100-point score across 8 signals: structured data, FAQ content, content clarity, E-E-A-T, citations, content depth, meta information, and internal linking. Includes specific recommendations for improvement. Example use: "Analyze https://sellonllm.com for AEO"
Check whether a website is cited/visible when an AI assistant (Claude, ChatGPT, Perplexity) answers a specific user query. Uses Perplexity's web-search AI to test the query and check if the target domain appears in the citations. Returns: visible (yes/no), citation position, context, and which competitors ranked instead. Example use: "Does https://sellonllm.com appear when someone asks 'best tools for selling on Amazon'?"
Compare your website's AI visibility against competitors for a set of queries. For each query, shows which sites get cited by AI and which don't — revealing where competitors are beating you and where you have an advantage. Example use: "Compare my site https://sellonllm.com against https://competitor1.com and https://competitor2.com for queries about 'selling on AI platforms'"
Automatically discover which user prompts/queries cause your website to appear in AI responses. Crawls your site to understand its content, generates relevant test prompts, then checks each one against AI search. Returns a ranked list of prompts where you're visible and a list of missed opportunities. Example use: "What prompts does https://sellonllm.com rank for in AI?" or "Find prompts for my e-commerce site"
Output schemas not documented. All tools return plain text strings via `{ content: [{ type: 'text', text: result }] }` rather than structured objects. LLMs cannot parse results or chain downstream tool calls.
Parameter constraints missing. 'max_pages' (1 - 20) and 'max_prompts' (default 10) lack min/max in descriptions. 'query' requires minLength 5 but this is not stated in description text. LLMs cannot validate inputs without explicit constraint documentation.
No error handling or recovery guidance. Tools do not document what happens on failure (invalid URL, API timeout, rate limit). LLMs have no guidance on retry strategy or fallback actions.
Inferred effective spec: 2026-07-28+.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | C | 60 | 2026-07-28+ | v2 |
Generate a comprehensive AI visibility report for a website across multiple queries. Optionally provide your own list of queries, or let the tool auto-generate them from your site content. Returns a full breakdown: visibility rate, which queries rank, missed opportunities, and recommendations. Example use: "Give me a full visibility report for https://sellonllm.com" or provide custom queries like ["how to sell on LLMs", "AI commerce tools"]
No idempotency or side-effect declarations. Tools call external APIs (Perplexity) and may have rate limits or costs, but descriptions do not warn about repeated calls or confirm-before-execute patterns.
Missing permission/scope declarations. Tools access external services and crawl websites but do not declare required permissions (read:website, call:perplexity_api) or rate limits.