Serenity Signal Dashboard

本项目抓取 x_curl/ 中的 X GraphQL curl，解析 @aleabitoreddit 的帖子、回复、订阅帖，抽取 $SYMBOL，写入本地 SQLite，并用 Yahoo chart 接口下载日线价格。

直接体验

如果你不想自己搭建本地环境，可以订阅 @iamai_omni，然后访问 app.k2ai.dev 直接使用托管版。也可以扫码直接打开订阅页：

本项目仅用于研究和可视化，不构成投资建议。

快速开始

python3 -m venv .venv
.venv/bin/python -m pip install -r requirements.txt

python3 scripts/ingest.py all --max-pages 10 --days 500 --min-mentions 3
python3 scripts/server.py --port 8787

打开 http://127.0.0.1:8787。

从 Chrome 复制 X curl

scripts/ingest.py fetch-x 会读取 x_curl/ 目录中的浏览器请求。首次使用或登录态过期时，需要从 Chrome DevTools 重新复制。

1. 用 Chrome 登录 X，并打开 https://x.com/aleabitoreddit。
2. 打开 DevTools：F12 或 Cmd/Ctrl + Shift + I。
3. 切到 Network 面板，筛选 Fetch/XHR，也可以在过滤框输入 UserTweets。
4. 刷新页面，滚动几次，触发帖子、回复或订阅内容加载。
5. 找到以下 GraphQL 请求，右键选择 Copy -> Copy as cURL。
6. 分别保存为这些文件名：

x_curl/UserTweets.curl
x_curl/UserTweetsAndReplies.curl
x_curl/UserSuperFollowTweets.curl

大致样例，真实内容会更长，并包含你的 cookie/token：

mkdir -p x_curl
cat > x_curl/UserTweets.curl <<'EOF_CURL'
curl 'https://x.com/i/api/graphql/.../UserTweets?variables=...&features=...' \
  -H 'authorization: Bearer ...' \
  -H 'cookie: auth_token=...; ct0=...' \
  -H 'x-csrf-token: ...' \
  -H 'x-twitter-active-user: yes'
EOF_CURL

注意：x_curl/*.curl 包含登录 cookie/token，已经被 .gitignore 忽略；不要提交或分享这些文件。

数据位置

• SQLite: data/serenity.sqlite
• 原始 X JSON: data/raw/*.json
• Dashboard: dashboard/index.html, dashboard/styles.css, dashboard/app.js

常用命令

python3 scripts/ingest.py fetch-x --max-pages 20
python3 scripts/ingest.py prices --days 700 --min-mentions 2
python3 scripts/ingest.py stats

注意：x_curl/*.curl 内的登录态可能过期；若抓取返回空或报错，重新从浏览器复制 curl 后再运行。

Codex Skill

本仓库内置了可开源分发的 Codex skill：skills/serenity-stock-scorer。它会基于本地 Serenity SQLite 快照，对单个 ticker 输出 0-100 的 Serenity 语料信号分。

python3 skills/serenity-stock-scorer/scripts/score_serenity_stock.py NVDA --pretty

该 skill 默认查找 data/serenity.sqlite；如果数据库在其他位置，可以传 --db /path/to/serenity.sqlite 或设置 SERENITY_DB_PATH。

---

Serenity Signal Dashboard (English)

This project reads X GraphQL curl commands from x_curl/, parses posts, replies, and premium posts from @aleabitoreddit, extracts $SYMBOL mentions, stores them in a local SQLite database, and downloads daily price bars from Yahoo's chart API.

Try It Directly

If you do not want to set up the local project yourself, subscribe to @iamai_omni, then visit app.k2ai.dev to use the hosted version directly. You can also scan this QR code to open the subscription page:

This project is for research and visualization only. It is not financial advice.

Quick Start

python3 -m venv .venv
.venv/bin/python -m pip install -r requirements.txt

python3 scripts/ingest.py all --max-pages 10 --days 500 --min-mentions 3
python3 scripts/server.py --port 8787

Open http://127.0.0.1:8787.

Copy X Requests From Chrome

scripts/ingest.py fetch-x reads browser-copied requests from x_curl/. You need to refresh these files when setting up the project or when the X session expires.

1. Log in to X with Chrome and open https://x.com/aleabitoreddit.
2. Open DevTools with F12 or Cmd/Ctrl + Shift + I.
3. Go to Network, select Fetch/XHR, and optionally filter by UserTweets.
4. Refresh the page and scroll a few times so X loads posts, replies, or premium content.
5. Find the GraphQL requests below, right-click each one, then choose Copy -> Copy as cURL.
6. Save them with these exact filenames:

x_curl/UserTweets.curl
x_curl/UserTweetsAndReplies.curl
x_curl/UserSuperFollowTweets.curl

Approximate example; the real command is longer and includes your cookie/token values:

mkdir -p x_curl
cat > x_curl/UserTweets.curl <<'EOF_CURL'
curl 'https://x.com/i/api/graphql/.../UserTweets?variables=...&features=...' \
  -H 'authorization: Bearer ...' \
  -H 'cookie: auth_token=...; ct0=...' \
  -H 'x-csrf-token: ...' \
  -H 'x-twitter-active-user: yes'
EOF_CURL

Warning: x_curl/*.curl contains login cookies/tokens and is ignored by .gitignore. Do not commit or share these files.

Data Files

• SQLite: data/serenity.sqlite
• Raw X JSON: data/raw/*.json
• Dashboard: dashboard/index.html, dashboard/styles.css, dashboard/app.js

Common Commands

python3 scripts/ingest.py fetch-x --max-pages 20
python3 scripts/ingest.py prices --days 700 --min-mentions 2
python3 scripts/ingest.py stats

If X fetching returns empty or invalid responses, copy fresh curl commands from Chrome and run the ingestion again.

Codex Skill

This repo includes an open-source-ready Codex skill at skills/serenity-stock-scorer. It scores a single ticker from the local Serenity SQLite snapshot as a 0-100 Serenity corpus signal.

python3 skills/serenity-stock-scorer/scripts/score_serenity_stock.py NVDA --pretty

The skill looks for data/serenity.sqlite by default. If your database lives elsewhere, pass --db /path/to/serenity.sqlite or set SERENITY_DB_PATH.