English doc below · 中文文档
A zero-dependency, local-first GitHub project-heat dashboard. It uses your already-logged-in gh CLI to pull Stars, Forks, clones, views and referrers for every repo under your account, and archives GitHub's 14-day-only Traffic data daily into a long-term trend.
Public stat tools (github-readme-stats / star-history / repobeats) cannot reach the owner-only Traffic API (clones / views / referrers) — they only rank Stars and Commits. This dashboard uses your local gh login session to pull that data directly: no token exposure, no server, just open index.html.
- Zero external dependencies — pure HTML/CSS with embedded data. No CDN, no build step, no backend.
- Real usage signals — clone count, view count, and traffic referrers.
- Long-term trend — GitHub keeps traffic for only 14 days; the script archives a daily snapshot into
history.json, so the trend line grows over time. - Self-contained single file —
collect.pyembeds the data intoindex.html; it works even fromfile://. - Any account — switch GitHub accounts with one env var or CLI argument.
# 1. Make sure gh is logged in (needs repo scope to read traffic)
gh auth status
# 2. Fetch and build the dashboard (defaults to truman-t3; override with arg / env)
python collect.py
# or: GITHUB_OWNER=your-name python collect.py
# or: python collect.py your-name
# 3. Open the dashboard
# Open index.html directly in a browser, or serve it:
python -m http.server 8777 # then visit http://127.0.0.1:8777/index.html- KPI bar — total Stars, 14-day clones, 14-day views, public repo count.
- Star vs Clones — grouped bars (neutral grey = Star, warm sand-gold = clone).
- Referrers — which channel is driving traffic (github.com / Google / Bing …).
- Long-term trend — daily snapshots of total clones + total Stars as line charts.
- Star history — cumulative Stars across a repo's lifetime (via the Stargazers API
starred_at). - Repo detail table — visibility, language, Stars, Forks, views, clones, unique clones.
GitHub only retains 14 days of traffic. Schedule collect.py to run daily and the trend line grows on its own:
# crontab example: run at 09:15 every day
15 9 * * * cd /path/to/github-heat-dashboard && /usr/bin/python3 collect.py(This project ships with a daily automation config inside WorkBuddy — enable it in one click from the chat.)
github-heat-dashboard/
├── index.template.html # page template (data placeholder /*__DASHBOARD_DATA__*/)
├── collect.py # fetch + archive + generate the single-file dashboard
├── gen_preview.py # render the cover image (preview.png) from real data
├── index.html # generated dashboard (overwritten each run, gitignored)
├── data.js # generated data file (gitignored)
├── history.json # daily snapshots (gitignored, feeds the trend)
├── preview.png # cover image
├── LICENSE
└── README.md
| Metric | Meaning |
|---|---|
| Star | Recognition / bookmark (a slow-moving vanity metric) |
| Fork | Someone wants to modify / reuse your code |
| Clone | The strongest "actually using it" signal |
| View | Exposure (14-day window) |
| Referrer | Where the traffic comes from (who is sending people) |
If this dashboard helped you see your GitHub reach at a glance, a star is the easiest way to say thanks — and it helps other developers find it. Click the ★ at the top-right of the repo.
MIT © truman-t3
一个零依赖、本地优先的 GitHub 项目热度看板。用已登录的 gh CLI 拉取你账号下所有仓库的 Star、Fork、克隆、浏览与流量来源，并把 GitHub 只保留 14 天的 Traffic 数据每日存档，画成长期趋势。
公开的 stat 工具（github-readme-stats / star-history / repobeats）都碰不到 owner-only 的 Traffic API（克隆 / 浏览 / 来源）——它们只做 Star / Commit 排行。本方案靠本机 gh 登录态直接拉取，不暴露 token、不搭服务器、双击 index.html 即开。
- 零外部依赖：纯 HTML/CSS + 内嵌数据，无 CDN、无构建、无后端
- 含真实使用信号：克隆量、浏览量、流量来源（referrers）
- 长期趋势：Traffic 仅 14 天窗口，脚本每日存档进
history.json，趋势曲线随时间增长 - 单文件自包含：
collect.py把数据内嵌进index.html，file://打开也能看 - 任意账号：改一个环境变量 / 命令行参数即可换 GitHub 账号
# 1. 确认已登录 gh（需 repo 权限以读取 traffic）
gh auth status
# 2. 拉取并生成看板（默认 truman-t3，可用参数 / 环境变量换账号）
python collect.py
# 或：GITHUB_OWNER=your-name python collect.py
# 或：python collect.py your-name
# 3. 打开看板
# 直接浏览器打开 index.html，或起本地服务：
python -m http.server 8777 # 然后访问 http://127.0.0.1:8777/index.html- KPI 数据栏：总 Star、近 14 天克隆、近 14 天浏览、公开仓库数
- Star 与克隆对比：分组双柱（中性灰 = Star，暖沙金 = 克隆）
- 流量来源：哪个渠道在带量（github.com / Google / Bing …）
- 长期趋势：每日快照累计的总克隆 + 总 Star 折线
- Star 历史：仓库整个生命周期的累计 Star（来自 Stargazers API 的
starred_at） - 仓库明细表：可见性、语言、Star、Fork、浏览、克隆、独特克隆
GitHub 自身只保留 14 天流量。建一个每日任务跑 collect.py，趋势曲线就会自己变长：
# crontab 例：每天 09:15 跑
15 9 * * * cd /path/to/github-heat-dashboard && /usr/bin/python3 collect.py（本项目在 WorkBuddy 中附带一个每日自动化配置，可在对话里一键启用。）
github-heat-dashboard/
├── index.template.html # 页面模板（数据占位符 /*__DASHBOARD_DATA__*/）
├── collect.py # 抓取 + 存档 + 生成单文件看板
├── gen_preview.py # 用真实数据渲染封面图（preview.png）
├── index.html # 生成的看板（每次运行覆盖，已 gitignore）
├── data.js # 生成的数据文件（已 gitignore）
├── history.json # 每日快照（已 gitignore，趋势数据源）
├── preview.png # 封面图
├── LICENSE
└── README.md
| 指标 | 含义 |
|---|---|
| Star | 认可度 / 书签（涨得慢的虚荣指标） |
| Fork | 想改 / 复用你的代码 |
| 克隆 | 最强「真要用」信号 |
| 浏览 | 曝光量（14 天窗口） |
| 来源 | 流量从哪来（谁在带量） |
如果这个看板帮你看清了自己在 GitHub 上的影响力，点个 Star 就是最简单的鼓励，也能帮更多开发者发现它。点击仓库右上角的 ★ 即可。
MIT © truman-t3