Enter keywords to search the latest works on Today's Toutiao (articles/videos), terminal table display + CSV export + interactive HTML report.
API Key must be configured first, passed via the environment variable REDFOX_API_KEY or the --api-key parameter. Search scope: Today's Toutiao platformlatest published works, fetched in real time, no cache delay.
Use cases
When you need to perform the following tasks, you should prioritize using this skill:
| Scenario | Example |
|---|---|
| Hot topic tracking | Search "AI" to learn about the latest AI-related content on Today's Toutiao |
| Viral content analysis | Search "large model" to find the viral articles with the highest read counts |
| Competitor content monitoring | Search brand keywords to track competitors' published content on Today's Toutiao |
| Topic inspiration collection | Search "new energy" and bulk export CSV for data analysis to find topics |
| Quick public opinion understanding | Search a certain event keyword to quickly understand the discussion trends on the Toutiao platform |
| Video content mining | Search "review" together with --video-only to filter viral video content |
Usage
# 基础搜索
python3 "$SKILL_PATH/scripts/search.py" "关键词"
# 按阅读量排序（默认按时间）
python3 "$SKILL_PATH/scripts/search.py" "AI" --sort views
# 限定最近 24 小时内发布的作品
python3 "$SKILL_PATH/scripts/search.py" "大模型" --hours 24
# 只查视频类作品
python3 "$SKILL_PATH/scripts/search.py" "测评" --video-only
# 多页翻页，获取更多结果
python3 "$SKILL_PATH/scripts/search.py" "新能源汽车" --pages 5
# 仅导出 CSV，不生成 HTML
python3 "$SKILL_PATH/scripts/search.py" "AI" --csv-only
# 不自动打开浏览器
python3 "$SKILL_PATH/scripts/search.py" "AI" --no-open
The terminal output is sorted inspecified sorting method(time/read count) in descending order; automatically prompts when there are few or no results.
HTML report features: dark theme · click the search button to query · click work cards to jump to the original text · article/video tags distinguished · paginated loading.
CSV / HTML are saved by default in ~/Downloads/QoderToutiaoSearch/。
Returned results display specification
When displaying search results to users,you mustuse the following complete field list, and must not omit any:
| No. | Field | Description |
|---|---|---|
| # | Ranking number | Starting from 1 |
| Title | Work title | Display at most 30 characters, truncate if exceeded |
| Author | Author nickname | — |
| Read count | viewCount | Abbreviated with k/10k (e.g. 3.2k, 12w) |
| Like count | likeCount | Same as above |
| Comment count | commentCount | Same as above |
| Share count | repostCount | Same as above |
| Publish time | publishTime | Format:YYYY-MM-DD HH:mm |
| Type | workType | Article / Video |
| Link | workUrl | Clickable to jump to the original text |
Table example:
| # | 标题 | 作者 | 阅读 | 点赞 | 评论 | 转发 | 发布时间 | 类型 |
|---|------|------|------|------|------|------|----------|------|
| 1 | 某某爆款文章标题... | 某某作者 | 3.2k | 412 | 89 | 56 | 2026-06-28 14:30 | 图文 |
Parameter description
| Parameter | Description | Default value |
|---|---|---|
keyword |
Search keywords (required, positional parameter) | — |
--sort |
Sorting method:views / time |
views |
--hours |
Limit to works published within the last N hours (0=unlimited) | 0 |
--pages |
Page-turn count (about 10 items per page) | 3 |
--video-only |
Return only video-type works | — |
--output-dir |
Output directory | ~/Downloads/QoderToutiaoSearch |
--api-key |
Specify API Key | — |
--no-open |
Do not automatically open the browser | — |
--csv-only |
Generate only CSV, not HTML | — |
--port |
HTML local service port | 8767 |
API Key configuration
Choose one way to configure your personal Key:
| 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 |
Registration address:redfox.hk
Features
- Dual-interface deep query：
searchWorkSearch list +workDetailRetrieve complete data (reads/likes/comments/reposts/shares) - Time filtering: Limit to fresh content within the last N hours, rejecting outdated cache
- Flexible sorting: Sort descending by publish time or read count
- Video filtering：
--video-onlyFind only video-type viral hits - Terminal table: Title, author, reads, likes, comments, reposts, publish time, link, type
- CSV export: Automatically generate UTF-8 BOM encoded CSV
- HTML interactive report: Dark theme, image/video tags distinguished, supports in-page search
- Local proxy service: Avoid browser cross-origin restrictions, support real-time in-page search
Dependencies
pip3 install requests
FAQ
Q: Can every viral hit be found? A: It can guarantee that each keyword search returns works.
Q: How far back does the search cover? A: Currently works from the past 1 year are provided, prioritizing recently published content.
Q: Why do I still have to wait for detail queries after searching? A: The search interface only returns basic information, while also providing per-work completion of full metrics such as read count, likes, and reposts, making the data more complete.
Q: How do I save the data after searching? A: You can have the AI export Excel for you, or have the AI export HTML.
Q: What should I do when the quota runs out? A: Go to redfox.hk to register and get a Token.
Q: What is the maximum number of items that can be retrieved?
A: About 10 items per page, 3 pages by default (about 30 items). It can be adjusted via the --pages parameter.