Language: English · 中文
Track what users of Claude Code, Codex, Cursor, and other AI coding products want by turning public GitHub issues into a bilingual demand observatory.
ReadYourUsers is a TypeScript workflow and public multi-product observatory for turning noisy GitHub issue streams into weekly demand maps across AI coding products:
- cross-product signal maps
- product-level deep dives
- rising demand shifts
- traceable issue evidence
- bilingual reports, compare pages, and product routes
Links: Live site · Compare · English site · 中文站点 · 中文版 README
95 issues analyzed from ruanyf/weekly · 1 need clusters · Updated 2026-04-06
| Rank | Need | Issues | Score | Category | Examples |
|---|---|---|---|---|---|
| 1 | AI Development Assistance and CLI Tooling | 31 | 1.6 | Developer Experience | #9519, #9511, #9507 |
| Need | Rising Score | This Week | Category |
|---|---|---|---|
| AI Development Assistance and CLI Tooling | 5.6x | 31 | Developer Experience |
Observatory | Compare | Product page | Full report | 中文版 | Based on public GitHub issues — signal, not census.
If you follow more than one AI coding product, reading GitHub issues repo by repo does not scale.
- titles are inconsistent across repos
- duplicates hide inside different product communities
- urgency is easy to miss in long queues
- cross-product shifts are hard to spot without a shared lens
ReadYourUsers turns that raw stream into a compact observatory you can scan in minutes, then drill into a specific product when something spikes.
- Cross-product observatory — homepage + compare view for tracked products
- Product deep dives — per-product reports and dedicated product pages
- Rising signals — what is accelerating right now
- Traceability — every insight links back to the original issues
- Bilingual outputs — English / Chinese reports and public site routes
- Repeatable workflow — fetch, analyze, aggregate, generate, publish
- Node.js 18+
- a GitHub token for reading public issues
- an OpenRouter key (recommended) or another OpenAI-compatible / Anthropic LLM key
npm install
cp .env.example .envREADYOURUSERS_GITHUB_TOKEN=your_github_token
LLM_PROVIDER=openai
OPENAI_API_KEY=your_openrouter_key
OPENAI_BASE_URL=https://openrouter.ai/api/v1
OPENROUTER_HTTP_REFERER=https://github.com/study8677/ReadYourUsers
OPENROUTER_APP_TITLE=ReadYourUsers
ANALYSIS_MODEL=qwen/qwen3.6-plus:free
AGGREGATION_MODEL=qwen/qwen3.6-plus:freenpx tsx src/cli.ts run
npm run site:buildnpx tsx src/cli.ts run anthropics/claude-code- Fetch — pull public issues from GitHub with caching
- Analyze — use an LLM to extract structured need signals
- Aggregate — cluster similar needs and compute demand / rising scores
- Generate — publish per-product reports, README snapshots, cross-product summary artifacts, and compare/product site routes
reports/latest/<repo>.md
reports/latest/<repo>.zh.md
reports/latest/cross-product.json
reports/archive/<week>/<repo>.md
reports/archive/<week>/<repo>.zh.md
site/en/index.html
site/en/compare/index.html
site/en/products/<slug>.html
site/zh/index.html
site/zh/compare/index.html
site/zh/products/<slug>.html
README.md
README.zh.md
| Repository | Product | Category |
|---|---|---|
| anthropics/claude-code | Claude Code | AI Coding CLI |
| openai/codex | OpenAI Codex CLI | AI Coding Agent |
| cursor/cursor | Cursor | AI Code Editor |
| opencode-ai/opencode | OpenCode | AI Coding Agent |
| cline/cline | Cline | IDE Coding Agent |
| Aider-AI/aider | aider | Terminal Pair Programming |
| block/goose | Goose | Open-source AI Agent |
| openclaw/openclaw | OpenClaw | Cross-platform AI Agent |
| google-gemini/gemini-cli | Gemini CLI | AI Coding Agent |
| RooCodeInc/Roo-Code | Roo Code | AI Coding Agent |
| continuedev/continue | Continue | AI Coding Assistant |
| microsoft/vscode-copilot-release | GitHub Copilot Chat for VS Code | AI Coding Assistant |
| voideditor/void | Void | AI Code Editor |
| zed-industries/zed | Zed | AI Code Editor |
| OpenHands/OpenHands | OpenHands | AI-Driven Development |
| TabbyML/tabby | Tabby | Self-hosted AI Coding Assistant |
- Public data only — private support channels are not included
- Signal, not census — issue volume is not the same as total user count
- LLM summarization is imperfect — conclusions stay grounded through source links
- Cross-product comparisons require care — GitHub issue culture varies by product
npm run build
npm run site:build
npm test- LINUX DO — 真正的技术社区
MIT