Why Doesn’t ChatGPT Know About a Product With 31,000 GitHub Stars?

Why don't GitHub stars translate into AI visibility?

• Stars are not text. A star count tells people a project is popular. It gives a language model nothing to learn about what the product does or when to use it.
• Training data lags. Models are trained on snapshots of the web, and research on knowledge cutoffs has found the effective cutoff often trails the stated one. Rapid recent growth may be invisible to answers from memory.
• Search needs the right pages. When ChatGPT searches, it needs pages that answer the buyer's question, such as "best open-source vector database for small teams". If comparison articles and tutorials do not include the project, it will not be named.
• Names get lost. A generic name, a name shared with other projects, or a repository name that differs from the product name makes it hard for systems to connect mentions.
• Documentation may be unreadable. Documentation sites that load content with JavaScript show most AI crawlers little text.
• Stars can be inflated. Researchers published "Six Million (Suspected) Fake Stars in GitHub", presented at ICSE 2026. It found fake-star activity surged in 2024, much of it promoting short-lived malware repositories and much of the rest AI, blockchain and tool repositories. Stars alone are weak evidence of quality, and they give an assistant no text to learn from.

Can ChatGPT read GitHub at all?

Yes. GitHub's robots.txt allows OpenAI's search crawler and its training crawler to access public repository pages. So a README can be read and cited when ChatGPT searches for it. But a README describes the project to people who have already found it. Buyers' questions about which tool to choose are usually answered from comparison and review content elsewhere.

How do you diagnose the gap?

1. Ask directly: "What is [project]?" and "What is [project] used for?" If the answer is wrong or empty, the model lacks basic information.
2. Ask category questions: "Best open-source [category] for [use case]" and "Alternatives to [competitor]". Note which projects are named and which sources are cited.
3. Check the cited sources for your project. Are you in the comparison articles, lists and tutorials that are cited?
4. Check your documentation site: view the page source of key pages for the text of your overview and features.
5. Check naming: does the web refer to your project consistently by one name?

How can open-source and developer products become known to AI?

• Write a clear "what is it" page on your site and at the top of your README. Say what it is, what problem it solves, who it is for and how it compares with alternatives.
• Make documentation readable to crawlers, with server-rendered or static pages.
• Publish comparison and alternatives pages that place your project in its category honestly.
• Get into the lists and comparisons that are cited, such as curated lists, comparison articles and "best tools" roundups in your category.
• Publish tutorials and examples for the use cases people ask about, and encourage community write-ups.
• Keep package registry pages complete, such as your PyPI or npm listing, with the same description and links.
• Launch and discuss in developer communities where the discussions are public and indexed, following each community's rules.
• Keep one name, and state former names clearly if you rename.

Getting a B2B software product recommended more broadly is covered in how to get a B2B software product recommended by AI assistants. How Hacker News launches can help is covered in how to get your brand mentioned on Hacker News for AI citations.

What might an illustrative case look like?

Consider an illustrative open-source workflow automation project with 31,000 stars, most gained in the past year. Asked "What is [project]?", ChatGPT without search says it has no information. With search, it names three older competitors for "open-source workflow automation for developers" and cites two comparison articles and a vendor blog. The project's documentation is a single-page app, and its README opens with installation commands. The maintainers add a server-rendered overview page and rewrite the README's first paragraph to state what the tool does and who it is for. They publish comparisons with the three competitors, and contributors write tutorials for common use cases. The project gets added to one of the cited comparison articles. Within two months, search-enabled ChatGPT answers name it for several workflow questions.

What do people ask about developer tools and AI visibility?

Will ChatGPT learn about our project in the next model update?

Possibly, if enough text about it exists on the web when the next training data is collected. Search-enabled answers can reflect it sooner.

Do more stars help at all?

Indirectly. Popular projects attract articles, tutorials and discussions, which are what assistants use.

Should we add an llms.txt file to our docs?

It is optional. Some developer tools read it, but it does not replace readable documentation and third-party coverage.

Does a Wikipedia article matter for developer tools?

It can help established projects with independent coverage. New projects rarely meet Wikipedia's notability rules.