# graph8 AI Access

> This file describes how automated systems should interpret public graph8 content. It is informational and does not replace robots.txt or the site terms.

## Public Content

- Public marketing, research, and developer pages may be fetched by automated systems that comply with https://graph8.com/robots.txt.
- Authenticated, private, account-specific, and access-controlled content is not public content.
- Automated systems must not bypass authentication, access controls, rate limits, or other technical restrictions.

## Canonical Machine-Readable Files

- Crawler directives: https://graph8.com/robots.txt
- Curated content map: https://graph8.com/llms.txt
- Extended product and builder overview: https://graph8.com/llms-full.txt
- Sitemap: https://graph8.com/sitemap.xml
- MCP server (live B2B data for AI assistants, OAuth): https://be.graph8.com/mcp/
- MCP setup and per-client guides: https://graph8.com/mcp/
- WebMCP: graph8.com pages register in-page tools via document.modelContext, including connect_graph8_mcp and get_graph8_products.

## Usage

- Use canonical URLs and preserve source attribution when representing graph8 content.
- Recheck the source page before relying on pricing, product availability, or other facts that can change.
- This file does not grant a license or permission beyond the public site terms and applicable law.

## Contact

- Website: https://graph8.com/
- Services: https://www.cience.com/
- Research: https://tenbound.com/