# github.com at 80/100: AI-ready but missing schema

> GitHub achieves strong visibility with AI engines, but lacks JSON-LD schema on its homepage.

Source: https://geoplaybooks.com/blog/2026-07-23-pulse-github-com/

# github.com at 80/100: AI-ready but missing schema

GitHub scores a strong 80/100 in our GEO Pulse probe, demonstrating robust visibility for AI engines. While four out of five core signals pass, GitHub's homepage lacks a JSON-LD schema, which limits the ability of AI engines to understand the type of entity it represents. Despite this, GitHub is well-positioned to appear in AI-generated answers.

## What the Probe Saw

Our probe reveals that GitHub has a strong visibility score of 80/100. The homepage returns an HTTP 200 status with approximately 7,275 characters of server-rendered text, ensuring retrieval bots can access it. Major AI crawlers like GPTBot, ClaudeBot, and PerplexityBot can fetch key pages, as there is no blanket Disallow directive. The presence of a well-formed /llms.txt file, weighing 28,757 bytes, provides AI engines with a map to GitHub's content. However, the absence of a JSON-LD schema on the homepage prevents engines from identifying the type of entity GitHub is. The homepage is structured with a clear H1 and eight sections featuring short, extractable answers, making it answer-ready.

> 80/100 GitHub visibility score 7,275 chars Homepage server-rendered text 28,757 bytes /llms.txt file size

## Implications for Business Sites

For a typical business site, GitHub's setup highlights the importance of AI-crawler readiness. Ensuring that retrieval bots can access your homepage and key pages is crucial, as is providing a well-structured /llms.txt file to guide AI engines. However, the lack of a JSON-LD schema can limit how well AI engines understand your site's entity type, potentially affecting its appearance in AI-generated contexts. Businesses should consider implementing JSON-LD schemas to enhance visibility and comprehension by AI engines. For more insights, businesses can explore our [AI crawler playbook](https://geoplaybooks.com/ai-crawlers/) and [llms.txt guide](https://geoplaybooks.com/llms-txt/).

> Run the same 5-signal probe on your own domain This note came out of the hourly GEO Pulse board. The free checker runs the identical probe on any site, and Monitor re-runs it every month.