
MaxAEO AI Visibility MCP
io.github.maxaeoPublished by GitHub account maxaeov0.4.0Updated Oct 24, 2026
Local-first MCP server for AI visibility, GEO/AEO, llms.txt, and AI crawler readiness audits.
Overview
A local MCP server that audits a public website's readiness for AI search, checking llms.txt, robots, sitemap, metadata and AI crawler access.
- What you can do
- Validate llms.txt and its linked URLs, robots and sitemap alignment for a site, with optional locale and market settings for output language and crawler list (global or China). Check AI crawler readiness: robots rules, sitemap availability, homepage metadata, canonical, schema and access for named AI and search crawlers. Build a combined local visibility report with a confidence score, a technical foundation score, top issues and a 7-day action plan.
- When to use it
- Use it when you want a quick, one-off check of whether a site is crawlable and understandable for AI search and answer engines, or when validating llms.txt, robots and structured data before publishing. It suits GEO/AEO readiness reviews run from inside a coding agent or chat client.
- Requirements
- Runs locally over stdio, typically via npx from the npm package; Node.js is needed. No account, API key or authentication is declared. It fetches the target site's public pages and files, so network access is required. Optional environment variables set locale, market and call-to-action URLs.
Installation
In SourceWeft
- Open MaxAEO AI Visibility MCP in the dashboard and add it to a workspace.
- Enable the server for the chats that should use its tools.
Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.
Other MCP clients
Follow the launch instructions in the repository.
README
MaxAEO AI Visibility MCP Server for GEO, AEO, and AI SEO
Local-first Model Context Protocol (MCP) server for AI visibility audits, GEO, AEO, AI SEO, llms.txt, and AI crawler readiness in Claude, Codex, Cursor, Windsurf, and other MCP-compatible agents.
It helps Claude, Codex, Cursor, and other MCP-compatible agents check whether a public website is crawlable, understandable, and ready for AI search workflows. By default, it only performs local/public-web checks. It does not call MaxAEO cloud APIs, upload user domains, invoke LLM APIs, or use hidden telemetry.
60-Second Setup
Install from npm:
Add to an MCP client:
Run an audit from your agent:
For command-mode Claude and Codex workflows, see the MaxAEO AI Visibility Agent Kit.
Directory Listing Copy
If you want to list this project in an MCP directory, awesome list, AI SEO resource list, GEO resource list, or llms.txt tooling index, use concise factual copy:
For Claude and Codex command workflows, list the MaxAEO AI Visibility Agent Kit.
Use Cases
- Run an AI visibility audit from an MCP client or coding agent.
- Check GEO / generative engine optimization readiness.
- Check AEO / answer engine optimization basics for answer engines and AI search.
- Validate
llms.txt, robots.txt, sitemap alignment, canonical URLs, noindex, metadata, and JSON-LD. - Generate a local 7-day action plan without using paid LLM or search APIs.
Common search intents this MCP server targets:
- AI visibility MCP server
- Claude SEO MCP
- Codex AI visibility audit
- GEO / AEO local website audit
llms.txtchecker- AI crawler readiness for GPTBot, ClaudeBot, PerplexityBot, Google-Extended, and AI search workflows
Tools
All tools support:
Crawler coverage by market
audit_ai_crawler_readiness selects its crawler list from market.
No Chinese AI product publishes crawler documentation, and none of them split
training and retrieval into separate user-agents the way GPTBot and
OAI-SearchBot do. Their answers are retrieved through the parent company's
existing search index, so the traditional search crawlers are what actually
gate citation domestically.
Every token in both lists is either documented by its operator or observed
across multiple independent domains' robots.txt. Unverifiable tokens that
circulate online — BaiduBot, DoubaoBot, TongyiBot, Kimi-User and
similar — are deliberately excluded, and a test enforces that. Auditing against
a user-agent that does not exist produces a finding nobody can act on.
Install
Run
Claude Desktop
Cost And Privacy
- No MaxAEO internal service calls by default.
- No LLM API calls by default.
- No domain uploads.
- No hidden telemetry.
- No site file modifications.
- Reports include a transparent MaxAEO CTA so users can continue into hosted monitoring when they want it.
Score Meaning
The top-level score is a local-only AI visibility confidence score, capped at 85/100 because this free local audit does not measure live AI engine recommendations, brand mentions, citation quality, sentiment, competitor share of voice, or historical trends.
technicalScore can still reach 100/100 when crawlability, llms.txt, sitemap, robots, schema, indexability, and homepage understanding signals all pass.
Example Output
Human-facing reports should render cta.markdown as a text link instead of printing the raw URL on its own line.
Test Cases
See examples/test-cases.md for public test cases covering a healthy foundation, missing llms.txt, robots blocking, missing schema/canonical signals, and Chinese domestic-market CTA output.
Contributing
See CONTRIBUTING.md for local-first scope, claim boundaries, and directory listing guidance.
Example Prompt
Chinese / domestic market:
Command-style agent prompts.
Claude:
Codex:
Locale And CTA Configuration
You can set locale and market per tool call, or use environment variables:
Priority:
- tool-call arguments
- environment variables
- defaults:
en-US+global
Upgrade Path
This MCP server is best for a fast one-time check inside your agent. For a better product experience, use the MaxAEO web app: interactive reports, saved history, continuous monitoring, brand tracking, competitor tracking, and shareable reports.
License
MIT
Source: README.md at commit 1ecc48b
Tools
0Version history
1- v0.4.0LatestSep 16, 2026
Score
50% of the applicable checks met
How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.
- Reached in the last 30 days25 pts
Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.
- Auth declared10 pts
Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.
- Tools listed15 pts
Answer tools/list without a credential, or ship a package that starts in a sandbox without one.
- DescriptionRequired15 pts
- Logo10 pts
- Documentation15 pts
- License10 pts