1. Introduction
One-Sentence Positioning: The all-internet aggregated hot topic TOP10 ranking, based on hourly collected hot topic data from 7 major platforms, outputs the TOP10 hot topic events with the highest comprehensive popularity through intelligent event recognition and cross-platform merging.
Core Value: Solves three major pain points for content creators and marketing operators in hot topic tracking:
- Hot topics are scattered and difficult to integrate: No need to check each platform one by one; aggregate the hot rankings of 7 major platforms at once
- Cross-platform comparison is difficult: Automatically identifies discussion differences and popularity performance of the same event across different platforms
- Trend judgment is vague: Intelligently predicts hot topic trends based on dimensions such as popularity value, time on the ranking, and platform coverage
Target Users: Content creators, marketing operations personnel, media editors, brand planners, and data analysts.
Not Supported: This skill does not support querying details of specific hot words; it only provides aggregated queries for the all-internet hot topic ranking.
2. Features
Core Features
| Feature Module | Capability Description | Core Value |
|---|---|---|
| 🔍 All-Internet Hot Ranking Aggregation | Real-time capture of hot search data from 7 major platforms | Get all-internet hot topics with one click, no more checking platform by platform |
| 🔗 Cross-Platform Event Recognition | Intelligently identifies expressions of the same event across different platforms | Automatically merges similar topics to avoid duplicate counting |
| 📊 Popularity Trend Prediction | Predicts trends based on popularity value, duration, and platform coverage | Judge the lifecycle of hot topics in advance and seize the creation window |
| 📈 TOP10 Ranking Provision | Outputs TOP10 hot topics sorted by comprehensive popularity | Quickly locate high-value topics |
| 💬 Cross-Platform Discussion Analysis | Displays discussion focuses and differences across different platforms | Gain deep insight into the public opinion ecosystem and precisely target audiences |
| 📄 HTML Report Export | Generates beautiful visual reports | Supports image export, convenient for sharing and archiving |
| ⏰ Subscription Push Service | Scheduled push of the latest hot ranking/yesterday's hot ranking | Continuously track hot topic dynamics and don't miss key opportunities |
Special Highlights
- Intelligent Event Recognition: Independently identifies and summarizes specific hot topic events from all titles, without directly using the original titles
- Visualized HTML Report: Automatically generates beautiful HTML reports and supports PDF/image export
- Cross-Platform Discussion Comparison: Displays discussion focuses and differences of the same event across different platforms
- Popularity Trend Prediction: Comprehensively analyzes popularity value, time on the ranking, and ranking changes to output trend predictions
3. One-Click Installation
Authentication
Get API Key
Please go to Honghu Hub to get the API KEY
Configure API Key
Option 1: Using OpenClaw as an example, add REDFOX_API_KEY to ~/.openclaw/openclaw.json, with part of the content as follows:
{ "env": { "REDFOX_API_KEY": "ak_xxxx..." } }
Option 2: Terminal configuration:
export REDFOX_API_KEY="ak_xxxx..."
Dependency installation
pip install python-dateutil==2.8.2
mkdir -p output
Environment variable configuration
| Variable name | Description | Required |
|---|---|---|
REDFOX_API_KEY |
Red Fox API Key | Yes |
4. Usage Guide
Basic Usage
Query the latest trending list (default)
python scripts/fetch_hotspot.py --output json
Automatically calculate the previous hour of the current time as the query range. For example: if the current time is 2026-04-16 08:30:00, then query data from 2026-04-16 07:00:00 to 2026-04-16 08:00:00 .
Query historical trending lists
# 查询昨日热榜（假设今天是2026-04-16）
python scripts/fetch_hotspot.py --start-date "2026-04-15 00:00:00" --end-date "2026-04-16 00:00:00"
Date range rules：
- The time format is
YYYY-MM-DD HH:MM:SS, and it can also be abbreviated asYYYY-MM-DD(automatically completed to 00:00:00) - The date range is
[start_date, end_date)a left-closed, right-open interval - Maximum query range:7 days
Advanced Usage
Standard execution process
Important principle: after the agent completes data analysis and saves the JSON, directly generate the HTML report, and do not output the detailed ranking list in the conversation.
- Obtain hotspot data: call the script to obtain the raw JSON data
- The agent analyzes and saves the JSON: perform hotspot event identification, sort in descending order by popularity value and take the TOP10, save as
structured_report.json - Automatically generate the HTML report：
python scripts/generate_html_report.py --input structured_report.json --output 热点榜报告.html - Output brief information in the conversation: title > statistics time > HTML report > subscription push service inquiry
Agent time judgment logic
| User intent | Query method |
|---|---|
| "Today's trending list" / "Today's hotspots" | Query from today's 0:00 to the current hour:--start-date "T 00:00:00" --end-date "T {当前小时}:00:00" |
| "Latest trending list" / "Hotspot list" | Query the previous hour of the current time:--start-date "T {当前小时-1}:00:00" --end-date "T {当前小时}:00:00" |
| "Yesterday's trending list" / "Yesterday's trending list" | --start-date "T-1 00:00:00" --end-date "T 00:00:00" |
| "Last 7 days trending list" / "Weekly trending list" | --start-date "T-7 00:00:00" --end-date "T 00:00:00" |
| "X month X day trending list" | --start-date "X月X日 00:00:00" --end-date "X月X日+1天 00:00:00" |
Comparison query: data for multiple days must be queried separately and cannot be combined into one query:
# 对比昨天和今天的热榜（假设今天是2026-04-16）
python scripts/fetch_hotspot.py --start-date "2026-04-15 00:00:00" --end-date "2026-04-16 00:00:00" # 昨日
python scripts/fetch_hotspot.py --start-date "2026-04-16 00:00:00" --end-date "2026-04-17 00:00:00" # 今日实时
Hotspot event identification rules
Core principle: completely ignore the keyword and category returned by the API, and independently identify and summarize specific hotspot events from all titles.
Identification process (must be executed in order):
- Collect all titles: traverse the
hotspotsarray, extracttitleandplatName - Identify specific events: determine whether the titles describe the same event (same subject, same event core, temporal continuity)
- Summarize event hot words: extract concise hot words for each event (5-15 characters)
- ✅ Correct: U20 Women's Football China-Japan Match, 2026 University Rankings Released
- ❌ Incorrect: China-related hotspots, sports news
- Sort by popularity value: take TOP10
Popularity value processing and output specifications
- Popularity conversion：
maxHotScore // 10000(integer division), append "万". For example: 9384468 → 938万 - Popularity value format: must be "number + 万", and other characters are prohibited
- Duration：
topOfTheDayTimeWhen it is 0, display "just made the trending list"; otherwise display "{N}h" - URL link: if there is a
url, it is displayed as a hyperlink; if there is no URL, only the text is displayed - Platform icon: distinguished using emoji (Weibo 🌐, Douyin 🎵, Zhihu 📚, Bilibili 📺, Kuaishou 🎬, Toutiao 📰, Baidu 🔍)
Comprehensive prediction rules
| Popularity range | Prediction emoji | Description |
|---|---|---|
| ≥ 10 million | 🔥🔥🔥 | Viral hit |
| 5-9.99 million | 🔥🔥 | High heat |
| 1-4.99 million | 🔥 | Medium |
| < 1 million | 📉 | Low heat |
The predicted content must be no less than 30 characters and must be comprehensively analyzed based on the topic type (breaking events/entertainment gossip/social livelihood/industry trends), popularity value, time on the list, and platform coverage performance.
Structured report JSON format
After the agent completes the analysis, save it as structured_report.json：
{
"query_range": { "start_date": "...", "end_date": "..." },
"hotspots": [
{
"rank": 1,
"title": "归纳的事件热词",
"hot_score": "938万",
"platform_count": 4,
"duration": "14h",
"max_position": 3,
"platforms": ["微博", "抖音", "头条", "快手"],
"discussions": [
{
"platform": "微博",
"focus": "讨论焦点描述，不少于10个字",
"topics": [{"title": "原始标题1", "url": "https://..."}]
}
],
"prediction": "预测内容文字",
"prediction_emoji": "🔥🔥🔥"
}
]
}
Key rules：
discussionsMust coverplatformsall platforms currently on the list inhot_scoreMust be in "number + ten thousand" format- The discussion focus for each platform must be no less than 10 characters
- Each platform displays 2-3 specific topic titles
Command quick reference table
| Scenario | Command |
|---|---|
| Latest hot list | python scripts/fetch_hotspot.py --output json |
| Today's hot list | python scripts/fetch_hotspot.py --start-date "T 00:00:00" --end-date "T HH:00:00" --output json |
| Yesterday's hot list | python scripts/fetch_hotspot.py --start-date "T-1 00:00:00" --end-date "T 00:00:00" --output json |
| Generate HTML report | python scripts/generate_html_report.py --input structured_report.json --output 热点榜报告.html |
Data structure description
{
"status": "success",
"stat_time": "2026-04-16 08:30:00",
"query_range": { "type": "realtime", "start_date": "...", "end_date": "..." },
"total_count": 50,
"hotspots": [
{
"hotId": "0DFEC...",
"title": "匈牙利总理用三个最描述中国",
"platName": "头条",
"plat": 11,
"url": "https://www.toutiao.com/trending/...",
"firstRankTime": "2026-04-15 21:00:00",
"latestRankDate": "2026-04-16 00:00:00",
"maxHotScore": 4427099,
"maxPosition": 15,
"topOfTheDayTime": "3",
"source_keyword": "中国"
}
]
}
| Field | Meaning | Analyzable dimensions |
|---|---|---|
| hotId | Hotspot unique ID | - |
| title | Hotspot title | Event identification, cross-platform merging |
| platName | Platform name | Platform coverage analysis |
| plat | Platform code | - |
| url | Redirect link | View details, redirect visit |
| firstRankTime | First time on the list | Hotspot fermentation starting point, timeliness |
| latestRankDate | Latest date on the list | Whether the hotspot is still on the list |
| maxHotScore | Highest popularity value | Popularity ranking, popularity comparison |
| maxPosition | Highest ranking position | Popularity peak, ranking changes |
| topOfTheDayTime | Time spent on the list (hours) | Popularity persistence, lifecycle prediction |
| source_keyword | Interface grouping keywords | For reference only, not for output |
Analyzable dimensions：
- Stay < 3 hours: short-term hotspot, rapid decay
- Stay 3-10 hours: medium persistence
- Stay > 10 hours: long-term hotspot, continuous fermentation
5. Use scenarios
Scenario 1: Topic selection decisions for content creators
Role: short video/self-media creator Needs: Quickly understand the 10 hottest topics across the internet every morning and judge which one is worth creating How to use: Enter "Today's hot topics" to get the TOP10 aggregated hot topics + HTML visualization report Expected benefits: Identify high-value topics within 5 minutes and increase content exposure
Scenario 2: Brand public opinion monitoring
Role: Brand public relations manager Needs: Quickly understand whether the current hottest events involve your own brand or competitors How to use: View the aggregated TOP10 across all platforms, and pay attention to differences in cross-platform discussions Expected benefits: Discover potential public opinion signals at the earliest opportunity and formulate response strategies in time
Scenario 3: Hotspot trend research
Role: Data analyst/researcher Needs: Analyze the recent evolution trends of hotspots and produce hotspot research reports How to use: Query the hot list data for the past 7 days and generate an HTML report for sharing and reporting Expected benefits: Data-based hotspot trend analysis to support decision-making and report writing
Scenario 4: Operational campaign planning
Role: Campaign operations planner Needs: Plan marketing campaigns by leveraging the current hottest topics How to use: View the TOP10 hotspots + heat trend forecasts, and choose hotspots in a rising phase to leverage Expected benefits: Precisely leverage hotspots to increase campaign participation and communication effectiveness
6. Project architecture
Directory structure
trending-hub-top10/
├── SKILL.md # 技能描述文件
├── scripts/
│ ├── fetch_hotspot.py # 热点数据获取脚本
│ └── generate_html_report.py # HTML报告生成脚本
├── references/
│ ├── output-templates.md # 输出格式模板参考
│ └── prediction-logic.md # 热度趋势预测规则
└── assets/
└── report-template.html # HTML报告模板
Technology stack
| Component | Technology | Description |
|---|---|---|
| Scripting language | Python 3 | Data acquisition and report generation |
| External dependency | python-dateutil | Date handling |
| Data interface | Redfox API | Multi-platform hotspot data aggregation |
| Report template | HTML/CSS/JS | Visualized HTML report |
| Output format | JSON / HTML | Structured data and visualized reports |
Core module description
| Module | Function |
|---|---|
fetch_hotspot.py |
Retrieve multi-platform hotspot data from the API, supporting time range queries |
generate_html_report.py |
Read structured_report.json to generate an HTML report (parameters: --input JSON path --output output path) |
output-templates.md |
HTML report format reference template |
prediction-logic.md |
Heat trend forecast rule reference |
report-template.html |
HTML report template, used to render the final report |
Resource index
- Script: see scripts/fetch_hotspot.py(Purpose and parameters: retrieve hotspot data from the API)
- Script: see scripts/generate_html_report.py(Purpose and parameters: read structured_report.json to generate an HTML report, parameters --input JSON path --output output path)
- Reference: see references/output-templates.md(When to read: HTML report format reference)
- Reference: see references/prediction-logic.md(When to read: refer to the forecast rules when generating heat trend forecasts)
- Asset: see assets/report-template.html(Used directly for generation: HTML report template)
7. FAQ
Installation-related
Q: The script reports an error when running: "ModuleNotFoundError: No module named 'dateutil'"
A: Please install the dependency:pip install python-dateutil==2.8.2
Q: Prompt "REDFOX_API_KEY not found"
A: Please ensure the environment variable has been configured REDFOX_API_KEY, and refer to the authentication section above for configuration.
Usage-related
Q: What is the basis for the TOP10 ranking? A: Sort in descending order by hotness value (maxHotScore), with the hottest ranked 1st. Before sorting, you must check one by one whether the hotness values are decreasing.
Q: Why do some trending topics appear on only one platform? A: This is normal. Different platforms have different user groups and content preferences, so some trending topics may only gain traction on specific platforms.
Q: Why can't I see the detailed rankings in the conversation? A: This skill is designed to output only brief information in the conversation (title, statistics time, subscription prompt), while the detailed content is displayed in the HTML report for easy sharing and exporting.
Q: How far back can data be queried? A: The maximum query range is 7 days.
Troubleshooting
Q: HTML report generation failed?
A: Please check: 1) structured_report.json whether it exists and is correctly formatted; 2) discussions whether it covers all platforms in platforms; 3) whether the hotness value format is correct ("number + wan").
Q: The hotness values in the report are inconsistent with the conversation?
A: Please ensure that the data in structured_report.json is completely consistent with the conversation output. The HTML report script is only responsible for template rendering and does not perform data analysis.
Q: Event identification is inaccurate? A: Trending event identification is performed by AI. If the identification is inaccurate, please try querying with a more specific time range, or use the trending-hub skill to view rankings categorized by platform.