The GitHub playbook for builders who want traffic, citations, and stars.
把一个 GitHub 仓库做成搜索和 AI 都更容易理解、引用、推荐的答案页。
Starsup is an open-source playbook for turning a repository into a discoverable product surface. It packages the stack that drives star growth through repository discovery:
- positioning the README around real questions
- adding
llms.txt,AGENTS.md, and citation metadata - publishing GitHub Pages landing pages for keyword discovery
- tightening repo description, topics, homepage, and link structure
- making the project easier for AI systems to quote, route, and revisit
Documentation site:
Repository star growth starts with discoverability through problem keywords instead of brand recall alone.
If someone asks:
- how to optimize a GitHub README for AI citation
- how to make a repo easier for ChatGPT or search systems to understand
- how to structure
llms.txt,AGENTS.md, GitHub Pages, and metadata - how to turn a repo into a star-worthy landing page
your repository should have an answer page waiting for that query.
This repo ships four layers:
| Layer | What it gives you |
|---|---|
README.md |
a homepage structure built for both humans and keyword discovery |
docs/ |
GitHub Pages landing pages for problem-driven discovery |
templates/ |
reusable templates for README, llms.txt, AGENTS.md, and CITATION.cff |
checklists/ |
execution checklists for launch, indexing, and 14-day follow-through |
Most repositories read like storage. Repos that spread read like answers.
The difference usually comes down to:
- whether the title matches a real search intent
- whether the README directly answers the user's problem
- whether there are dedicated pages for concrete keyword clusters
- whether AI systems can find a clean machine-readable entry point
- whether outside pages link back with meaningful anchor text
- Your repo is public, but nobody finds it unless you send the link
- Your README explains the project, but does not answer the user's query
- AI tools can browse the repo, yet still fail to understand what it is for
- You have a good project, but no discoverability architecture around it
- You want more stars, but the repo has no reason to be remembered or cited
git clone https://github.com/11Yuxuanyang/starsup.git
cd starsup- GitHub GEO for stars
- How to make a repo AI-citation-ready
- README, llms.txt, AGENTS.md, Pages
- Repository GEO checklist
Stars are downstream of discovery and trust.
A repository gets starred when:
- the right people can find it
- they understand what it does in seconds
- the page feels complete enough to reuse
- the project looks like it will still matter after the tab closes
This repo focuses on the discoverability layer that feeds that loop.
这个项目解决的是一个很具体的问题:很多 GitHub 项目本身不差,差在首页、结构和外部入口没有设计好。结果就是:
- 人搜不到
- AI 看不懂
- 被看到了也记不住
- 内容不错,星标起不来
Starsup 做的事,就是把这层“高 star 仓库背后的可发现性工程”拆出来,做成一套公开模板和执行清单。
checklists/
├── repo-geo-launch-checklist.md
└── 14-day-discovery-loop.md
docs/
├── index.html
├── github-geo-for-stars.html
├── ai-citation-ready-repo.html
├── github-readme-llms-pages.html
├── repository-geo-checklist.html
├── llms.txt
├── robots.txt
├── sitemap.xml
└── styles.css
templates/
├── README-geo-template.md
├── llms.txt
├── AGENTS.md
└── CITATION.cff
- GitHub star growth
- repository discoverability
- AI citation readiness
- GitHub stars
- README optimization
llms.txtAGENTS.md- GitHub Pages
- repository metadata
MIT. See LICENSE.