last30days-cn skill
A skill that gathers 30 days of discussion on eight Chinese platforms, ranks it and writes a cited report, and can also show the cross-platform hot lists
- 1,941 ★ on the repository
- Last commit
Broad access
Why this level
- Keeps login cookies for Weibo, Zhihu, Bilibili and Douyin and API keys on disk and then uses them in requests
- Reaches eight platforms and the paid services TikHub, jisuapi and Baidu Qianfan
- Runs Python scripts and opens a Chromium browser during login
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/last30days-cnDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- Keeps login cookies for Weibo, Zhihu, Bilibili and Douyin and API keys on disk and then uses them in requests
- Reaches eight platforms and the paid services TikHub, jisuapi and Baidu Qianfan
- Runs Python scripts and opens a Chromium browser during login
This is third-party code. Review the repository files before installing.
What it does
last30days-cn is a version of the last30days skill for the Chinese internet. When the agent asks, scripts/last30days.py queries Weibo, Xiaohongshu, Bilibili, Zhihu, Douyin, WeChat public accounts, Baidu and Toutiao in parallel. For each platform it tries paths in order: an API, a saved login, a public interface, search over open pages. Without a login, some platforms answer only with a hot list and open search, so the result has links without reliable dates or engagement figures. The skill then cleans what it found: it unifies the fields, drops items outside the search window, duplicates and off-topic entries, and keeps up to three per author. Each entry gets a score from 0 to 100: relevance 45%, recency 25%, engagement 30%. Baidu and WeChat are scored without engagement. An event covered on several platforms merges into one hot topic. The agent receives a compact digest, while the script writes the full reports report.md, report.json and report.html to ~/.local/share/last30days/out/. The end of the report labels each platform with its item count and data path, and failures come with a cause and a fix command. The --hot mode needs no topic: it pulls the hot lists of Weibo, Baidu, Douyin, Toutiao, Bilibili and Zhihu, merges matching topics and ranks higher those that reached more platforms. --hot-sources adds RSS news feeds and Hacker News. --global brings in Hacker News, GitHub and Reddit by the English query from --global-query.
Who it is for. For analysts, marketers and founders who need to see what the Chinese internet says about a product, event or topic and want the agent to read the Chinese sources for them.
Good fit when
- You want to know how Chinese users discuss a product, event or competitor over the last 30 days or a shorter window
- Before entering the Chinese market you want to compare the tone of platforms: experience on Xiaohongshu, analysis on Zhihu, reactions on Weibo, reviews on Bilibili
- You need a summary of what is on the hot lists of Weibo, Baidu, Douyin, Toutiao, Bilibili and Zhihu right now, with repeated topics grouped
- You need a report as an HTML file that opens offline, with a source link behind every conclusion
- Some platforms do not answer and you need to see which one is off and how to fix it
Not a fit when
- You need bulk collection of platform data or information about people: the author explicitly excludes such use
- The data is for a paid product or service: the author permits the project for study and research only
- You need exact engagement figures for Weibo, Xiaohongshu, Zhihu and Douyin but will not sign in to accounts: without a login the skill returns links and hot lists
- You want Reddit, X and YouTube discussion without Chinese platforms: the original last30days in the catalog is built around those sources
- The machine has Python older than 3.8: the scripts target 3.8 and newer
Example request
Use last30days-cn to research what the Chinese internet says about the electric car price war over the last 30 days and give me a report with linksLimitations
The license is MIT, with the text in the LICENSE file; GitHub does not detect it automatically. In its disclaimer the author limits use to study and research, asks you to follow platform terms and PRC law, not to collect data in bulk and not to store other people's personal details. The author advises at least 5 seconds between searches.
Search on Weibo, Xiaohongshu, Zhihu and Douyin usually needs a login. You sign in once with the login command: a Chromium window opens through Playwright and you scan a QR code yourself. Without a graphical shell you can paste cookies from your own browser. Logins are stored in ~/.config/last30days-cn/browser_cookies/ and readable only by your user.
For a Chinese topic the agent translates the English query for overseas sources itself. Reddit rejects some cloud addresses. X, YouTube, TikTok and Instagram go through the separately installed last30days skill, which needs Python 3.12 and its own keys.
This is a localized version of someone else's skill, last30days. Platforms change their interfaces without notice, and the author does not promise every path works at all times. Search results are cached for 24 hours; --refresh ignores the cache.
How to disable. Delete the skill folder, for example ~/.claude/skills/last30days-cn after a manual install. Saved logins, the .env settings file and reports live apart in ~/.config/last30days-cn/ and ~/.local/share/last30days/out/, so delete them by hand. Daily hot-list publishing to GitHub Pages turns on only with the HOT_PAGES=true variable in your own repository and never starts by itself.
FAQ
How does the skill score what it finds?
The overall score from 0 to 100 combines relevance (45%), recency (25%) and engagement (30%). Baidu and WeChat public accounts are scored without engagement. A cap of three items per author keeps one voice from flooding the result.
Which platforms work without a login?
Bilibili and Toutiao work fully. WeChat goes through Sogou search and Baidu through web search. Weibo, Xiaohongshu, Zhihu and Douyin without a login answer with hot lists and open search, meaning links without exact dates or engagement figures, and the report marks such results separately.
Where does the Douyin, WeChat and Baidu data come from?
Douyin uses the TikHub API, WeChat public accounts the jisuapi service API, and Baidu the Qianfan search API. The keys TIKHUB_API_KEY, WECHAT_API_KEY and BAIDU_API_KEY are optional: without them the skill falls back to other paths, which return less data.
How do I run it on an old computer?
The scripts use only the standard library of Python 3.8, and a browser is needed only for platforms with a login. If the Playwright Chromium will not start, point LAST30DAYS_BROWSER_PATH at the system browser or turn the browser off with --no-browser. A failed launch switches the run to browserless mode by itself.
Can I publish the hot list as a website?
The repository has the daily-hot.yml workflow for GitHub Actions. It publishes the hot list to GitHub Pages three times a day at Beijing time (08:00, 12:00, 20:00). It is off by default: you need the Pages source set to GitHub Actions and the HOT_PAGES=true variable.
What does --diagnose show?
The state of every platform: whether it works, which path it uses and how to fix a failure. With --probe-browser the script really launches the browser once, and with --emit json it returns the result in a form convenient for an agent.
Discussion
No questions or reviews yet.
Related
SkillsA skill that gathers the last 30 days of discussion on a topic from Reddit, X, YouTube, HN, Polymarket and GitHub into one brief
The news-extractor skill pulls the title, text, images and metadata from a link on 12 news platforms and returns JSON or Markdown, with no browser
A skill that reads a link and hands the agent the full text as Markdown, picking a separate route for WeChat pages, Feishu docs, PDFs and arXiv papers
CLI and skill that let an agent read and search Twitter, Reddit, YouTube, GitHub, Bilibili and Xiaohongshu without paid APIs
A skill that routes a question to the right AI platform and, via browser automation, collects the answer from its native search
A skill for web search, X search and page fetching for models without web access: no signup or key, the answer is JSON with source links
Comparison with similar tools
Entries that share the main search query, and how they differ.
| Entry | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| last30days-cn (this entry) | no | 1,941 | ||
| last30days | no | 65,041 | the original skill: discussion from Reddit, X, YouTube, Hacker News, Polymarket and GitHub | |
| xiaohongshu-mcp | no | not required | 16,449 | an MCP server that publishes and searches Xiaohongshu notes for the agent; last30days-cn can use it as one data path |
| New-term radar | no | 1,676 | three skills and an MCP server find new names of AI models |
README in short
Documentation comes in Chinese and English (README.md and README.en.md). The repository holds SKILL.md with the modes and output rules, SPEC.md and AGENTS.md for developers, the release-notes.md changelog, regression tests that need no network, and CI. Manifests cover Claude Code (.claude-plugin/plugin.json and marketplace.json, version 4.0.0), Gemini CLI (gemini-extension.json) and OpenAI (agents/openai.yaml). A SessionStart hook shows whether keys and logins are configured. A platform table lists the data paths and what works without setup.
README badge
Original SKILL.md
---
name: last30days-cn
version: "4.0.0-cn"
description: "Research what Chinese internet users actually said in the last 30 days across Weibo, Xiaohongshu (RED), Bilibili, Zhihu, Douyin, WeChat public accounts, Baidu and Toutiao: engagement-weighted ranking, cross-platform clustering, cited Markdown/JSON/HTML reports. Also shows today's cross-platform hot-search board (全网热榜, --hot) and can add Hacker News / GitHub / Reddit or an installed upstream last30days for X/YouTube/TikTok."
argument-hint: 'last30 AI 编程助手 | last30 今天全网热点 | last30 具身智能 --deep | last30 Claude Code 评测 --global'
allowed-tools: Bash, Read, Write, WebSearch
author: Jesse
license: MIT
user-invocable: true
metadata:
openclaw:
emoji: "📰"
requires:
optionalEnv:
- WEIBO_COOKIE
- ZHIHU_COOKIE
- BILIBILI_COOKIE
- TIKHUB_API_KEY
- WECHAT_API_KEY
- BAIDU_API_KEY
- WEIBO_ACCESS_TOKEN
- XIAOHONGSHU_API_BASE
- GITHUB_TOKEN
bins:
- python3
files:
- "scripts/*"
tags:
- research
- deep-research
- chinese-platforms
- trending
- hot-search
- weibo
- xiaohongshu
- bilibili
- zhihu
- douyin
- wechat
- baidu
- toutiao
- html-report
---
# last30days-cn
You are a Chinese-internet research assistant. Use this skill when the user wants recent Chinese-platform discussion, public sentiment, product reputation, trend research, "最近 30 天" coverage, or **what is trending right now** (热搜 / 热榜 / 今天有什么热点).
## Core rule
Ground every claim in the returned evidence. Never invent sources, links, engagement numbers, dates, quotes or platform sentiment. When coverage is thin or a platform failed, say so plainly.
## 1. Pick the mode
| The user asks… | Run |