SuperAPI AI Directory
Partner中文AI newsFavorites
Skillsks-ai-feed

ks-ai-feed

Marketing & growth

View repository

Updated 10-16 04:46

About this skill

AI Kuaishou Information Source

📝 Introduction

Automatically scans AI-created content on the Kuaishou platform daily, filters viral videos by play count, intelligently clusters them to generate a dark-themed HTML daily report, and simultaneously outputs AI intelligence insights based on the intelligent intelligence investigator methodology.

✨ Features

Feature Module Capability Description Core Value
Viral Discovery Filter Kuaishou AI trending videos by play count Precisely locate high-popularity content
Intelligent Clustering Automatically identify topic directions from content Daily hot topics at a glance
AI Intelligence Insights Top trending topics, growth signals, influencer analysis, recommended investigation directions In-depth insight into traffic trends
Visualized Daily Report Dark-themed HTML daily report, including cover images, engagement data, and direct video links Intuitive browsing and navigation
One-Click Subscription --subscribe Enable daily automatic output Continuous tracking without missing anything
Image Compatibility Anti-hotlink proxy + automatic HEIF/HEIC to JPG conversion Stable display of cover images

🔑 Authentication

Go to redfox.hk/login to register and obtain an API Token, then configure:

Method Command
Environment Variable (Recommended) export REDFOX_API_KEY=ak_你的密钥
Command-Line Parameter --api-key ak_你的密钥
Configuration File echo '{"api_key":"ak_你的密钥"}' > ~/.qoder/apis/redfox.json

🔌 API Interface Specification

Batch Query Interface

Interface Address: POST https://redfox.hk/story/api/parseWork/queryKsAiMsgs/batch

Request Headers:

X-API-Key: ak_你的API密钥
Content-Type: application/json

Request Body:

{
  "keywords": ["AI", "人工智能", "大模型", "GPT", "Agent", "AI绘画", "AI教程"],
  "pageNum": 1,
  "pageSize": 200,
  "source": "AI快手信息源-GitHub",
  "startTime": "2026-06-16 00:00:00",  // 可选
  "endTime": "2026-06-16 23:59:59"      // 可选
}

Parameter Description:

Parameter Type Required Description
keywords Array ✅ Keyword array, supports batch queries with multiple keywords
pageNum Number ✅ Page number, starting from 1
pageSize Number ✅ Number of items per page, default 200
source String ✅ Source identifier
startTime String ❌ Start time YYYY-MM-DD HH:mm:ss
endTime String ❌ End time YYYY-MM-DD HH:mm:ss

Performance Advantages:

  • ✅ Pass all keywords in a single request, reducing API calls by 85%+
  • ✅ Automatic deduplication and merging for more comprehensive data
  • ✅ 30-second timeout, suitable for large batches of data

📊 Data Update Rules

  • Update Time: Updates the previous day's data daily at 15:00
  • Queryable Dates: Only dates for which data has already been updated can be queried (yesterday and earlier)
  • When Data Is Unavailable: The script automatically determines whether data already exists for the target date; if it has not yet been updated, it prompts the user and asks whether to query the latest available date

🔧 Usage

# 生成今日爆款日报（含情报洞察）
python3 "$SKILL_PATH/assets/daily_report.py"

# 自定义关注方向
python3 "$SKILL_PATH/assets/daily_report.py" --keywords "AI教程,AI绘画,ChatGPT,AI工具"

# 查看历史某天
python3 "$SKILL_PATH/assets/daily_report.py" --date 2026-06-10

# 订阅 / 取消订阅
python3 "$SKILL_PATH/assets/daily_report.py" --subscribe
python3 "$SKILL_PATH/assets/daily_report.py" --unsubscribe

Dependencies:pip3 install requests

⚠️ Note for Windows Users: In the command, python3 needs to be replaced with python，pip3 replaced with pip。

The generated HTML daily report is saved in ~/Downloads/QoderReports/. The terminal also outputs a categorized video table + AI intelligence insights.

Preview service

  1. When running daily_report.py , add the --no-open parameter
  2. Start the preview service separately(run in background): 
    python3 "$SKILL_PATH/assets/preview_server.py"
  3. Call RunPreview,you must use the direct HTML file URL：http://127.0.0.1:8766/{HTML文件名}

⚠️ RunPreview must use the direct HTML file URL(e.g. http://127.0.0.1:8766/AI快手日报_2026-06-16_150715.html), the root path http://127.0.0.1:8766must not be used; the 302 redirect on the root path will result in a blank page.

The built-in service also provides: static HTML file service + /api/search search proxy + /api/img image proxy (bypassing Kuaishou's hotlink protection).

Core parameters

Parameter Description Default value
--keywords Topic directions to follow, comma-separated AI,人工智能,大模型,GPT,Agent,AI绘画,AI教程
--count Number of videos to scan 200
--date Specify date,YYYY-MM-DD or range YYYY-MM-DD~YYYY-MM-DD today
--output-dir Output directory ~/Downloads/QoderReports
--api-key Specify API Key —
--subscribe Enable daily subscription —
--unsubscribe Disable daily subscription —
--no-open Do not automatically open the browser —

🧠 AI intelligence insights

Based on the intelligent intelligence investigator methodology, generated in sync after daily clustering of Kuaishou viral content:

Insight module Content Value
Top trending topics Share, count, top videos Quickly grasp the day's trends
Emerging growth signals Potential topics with low share but high engagement Capture traffic trends in advance
Core creators High-output + high-engagement authors Lock onto key creators
Recommended investigation directions Scenario + engine combination + search keywords One-click entry into in-depth investigation
Cross-platform comparison suggestions Simultaneously follow the same topics on other platforms Eliminate bias with a whole-network perspective

For detailed investigation strategies, see engine-strategy.md、investigation-modes.md、investigation-templates.md。

📋 Fixed output mode

After each daily report run, the terminal and conversation outputmustfollow the following structured format:

## AI快手信息源 · {日期} 日报

**扫描 {N} 条热门视频，聚类 {M} 个分类**

---

### 分类概览

| 分类 | 数量 | 占比 | 亮点 |
|------|------|------|------|
| #{分类名} | {N}条 | {X}% | 头部视频亮点描述 |

---

### AI情报洞察报告

**一、新兴起量信号**

- 🔥 **#{话题}** — 仅{N}条但均互动{X}+，描述

**二、核心达人**

| 达人 | 作品数 | 总赞 | 亮点 |
|------|--------|------|------|
| @{作者} | {N}条 | {X}w | 描述 |

**三、推荐调查方向**

| 话题 | 调查场景 | 推荐引擎 | 搜索关键词 |
|------|---------|---------|-----------|
| #{话题} | {场景} | 引擎组合 | 关键词1 / 关键词2 |

**四、跨平台对比建议**

- **{话题}** — 建议同步关注抖音、B站、小红书同话题热度

---

**日报地址**：{HTML文件绝对路径}

**预览地址**：http://127.0.0.1:8766/{HTML文件名}

The above format is amandatory specification, and all fields must not be omitted. If a module has no data, mark it as "None".

Must execute: After the daily report output is complete, start the preview according to the following steps:

  1. Start the preview service in the background:python3 "$SKILL_PATH/assets/preview_server.py"
  2. Call RunPreview, using the direct HTML file URL http://127.0.0.1:8766/{HTML文件名}(the root path must not be used)

Other resources