llm-intelligent-public-opinion-analytics

Deploy and use an LLM-powered public opinion analytics assistant that crawls 26 hot lists from 15 platforms, performs sentiment analysis, topic clustering, and multi-channel alerting

3stars1forksUpdated 7/24/2026

Data Science & ML #web-scraping #monitoring #llm #python #data-analytics #sentiment-analysis

Security Assessment

Low Risk(80/100)

Detected risks:

Automated Scraping With Evasion Features(The skill configures a crawler cluster that scrapes 26 hot lists across 15 third-party platforms, including browser-driver scraping of detail pages., Crawler settings include user-agent rotation and configurable concurrency, which are patterns that can be used to evade anti-bot controls and may conflict with target platforms' terms of service., It supports supplying authenticated platform session cookies (for example for Weibo and Bilibili) to obtain better access to content behind login.)

Centralized Secrets And Outbound Webhooks(A single local .env file collects a large set of sensitive credentials including database passwords, an LLM API key, SMTP email credentials, WeChat Work corporate secrets, and a Telegram bot token., The system sends collected and analyzed content outward through multiple notification channels and webhooks, so misconfiguration could route data to unintended recipients.)

Security Score80/100

About llm-intelligent-public-opinion-analytics

This skill is a deployment and usage guide for an LLM-powered public opinion analytics platform (an upstream open-source project it points to) that crawls 26 hot lists across 15 mainstream Chinese platforms such as Weibo, Bilibili, Zhihu, and Baidu, then layers on sentiment analysis, topic clustering, and multi-channel alerting. It solves the problem of aggregating cross-platform trending content and turning it into queryable, analyzable intelligence with a conversational interface.

The skill documents end-to-end setup: installing a Chrome/Edge browser driver for detail-page scraping, provisioning a MySQL 8.0+ database, cloning the upstream repository, and configuring a `.env` file with database credentials, an OpenAI-compatible (or Huawei Pangu) LLM endpoint, and push-notification credentials for Email/SMTP, WeChat Work, and Telegram. Crawler settings cover a connection pool, optional platform cookies for authenticated access, user-agent rotation, and concurrency/delay controls. Usage covers starting a Flask web interface, managing crawlers, running natural-language queries, and calling a programmatic analyzer API for hot-search queries, sentiment scoring, and topic clustering.

Target users are analysts, communications and PR teams, and developers who need to monitor social-media trends and public sentiment at scale. Note that this is largely a catalogue-style pointer to an external repository, and its scraping behavior warrants care: it uses browser-driver scraping, user-agent rotation, and optional authenticated cookies against third-party platforms, plus it centralizes many credentials and outbound webhook/notification tokens.

FAQ

What infrastructure do I need to run this?

A Chrome or Edge browser driver on PATH for detail-page scraping, a MySQL 8.0+ database, a Python virtual environment with the project's requirements installed, and an OpenAI-compatible (or Huawei Pangu) LLM API endpoint.

Which platforms and notification channels are supported?

It aggregates 26 hot lists from 15 platforms including Weibo, Bilibili, Zhihu, and Baidu, and can push alerts via Email/SMTP, WeChat Work (bot or application), and Telegram.

Is the skill the tool itself or a guide to it?

It is primarily a deployment and usage guide that clones and configures an external upstream repository; the analytics platform code lives in that separate GitHub project.

Are there compliance or rate-limiting considerations when scraping?

Yes. The setup uses browser-driver scraping with user-agent rotation, configurable concurrency and download delay, and optional authenticated platform cookies. Users are responsible for respecting each platform's terms of service, rate limits, and applicable data/privacy laws.

How are credentials handled?

Database, LLM, SMTP, WeChat Work, and Telegram credentials are stored in a local .env file and read by the crawler settings; you supply and control these secrets, but the configuration centralizes many sensitive tokens in one place.

Install llm-intelligent-public-opinion-analytics

Download and extract the skill files to your .claude/skills/ directory.

Quick Setup:

Copy the skill folder to .claude/skills/
Claude will automatically detect and use the skill

Repository

aradotso/data-skills

Related Skills

short-form-video

14

multi-asset-morning-briefing

7,907

content-strategy

2,901

doherty-threshold

2,901