ไปๆฅไธญๅๅฎ่งๆฐ่ Skill
๐ Based on news-aggregator-skill | ๅฐๆณจๆผไธญๅๅฎ่ง็ถๆฟๆฐ่็ๅ็ดๆดๅฑ
ๅพๅคๅไธญๆ่ฒก็ถๆฐ่ๆบๆๅไธฆ็ฏฉ้ธไธญๅๅฎ่ง็ถๆฟ็ธ้ๆฐ่๏ผๆไพ AI ๆทฑๅบฆ่งฃ่ฎใ
Tools
fetch_china_macro_news.py
Usage:
python scripts/fetch_china_macro_news.py --source wallstreetcn --limit 15
python scripts/fetch_china_macro_news.py --source wallstreetcn,36kr --limit 10
python scripts/fetch_china_macro_news.py --source wallstreetcn --limit 10 --deep
ๆบๆ
ง้้ตๅญๆดๅฑ (Smart Keyword Expansion)
CRITICAL: ็ถ็จๆถ็ตฆๅบ็ฐกๅฎ้้ตๅญๆ๏ผ่ชๅๆดๅฑ่ฆ่็ธ้้ ๅ๏ผ
- ็จๆถ: "ๅฉ็" -> Agent ไฝฟ็จ:
--keyword "ๅฉ็,LPR,MLF,้ๆฏ,ๅ ๆฏ,PBOC,ๅคฎ่ก"
- ็จๆถ: "้่จ" -> Agent ไฝฟ็จ:
--keyword "้่จ,CPI,PPI,็ฉๅน,้็ธฎ"
- ็จๆถ: "่ฒฟๆ" -> Agent ไฝฟ็จ:
--keyword "่ฒฟๆ,้ฒๅบๅฃ,้ ๅทฎ,้็จ
,ๆตท้"
python scripts/fetch_china_macro_news.py --source wallstreetcn --limit 20 --keyword "ๅคฎ่ก,PBOC,ๅฉ็,LPR,MLF,้ๆฏ,้ๆบ" --deep
Arguments:
--source: One of wallstreetcn, 36kr, all (default: wallstreetcn).--limit: Max items per source (default 15).--keyword: Comma-separated filters (default: ๅฎ่ง็ธ้้้ตๅญ).--deep: [NEW] Enable deep fetching. Downloads and extracts the main text content of the articles.
Output:
JSON array. If --deep is used, items will contain a content field associated with the article text.
้ ่จญๅฎ่ง้้ตๅญ
่
ณๆฌ้ ่จญไฝฟ็จไปฅไธ้้ตๅญ็ฏฉ้ธไธญๅๅฎ่งๆฐ่๏ผ
ๅคฎ่ก,PBOC,ๅฉ็,LPR,MLF,้ๆฏ,้ๆบ,
GDP,PMI,CPI,PPI,้่จ,้็ธฎ,
็ถๆฟ,ๅฎ่ง,่ฒกๆฟ,่ฒจๅนฃๆฟ็ญ,
่ฒฟๆ,้ฒๅบๅฃ,้ ๅทฎ,้ๅทฎ,
ๅฐฑๆฅญ,ๅคฑๆฅญ,ๆถ่ฒป,้ถๅฎ,
ๆฟๅฐ็ข,ๆจๅธ,ๆ่ณ,ๅบๅปบ,
ไบบๆฐๅนฃ,ๅฏ็,ๅคๅฏ,
ๅตๅธ,ๅๅต,ไฟก่ฒธ,็คพ่,M2
Interactive Menu
When the user says "ไปๆฅไธญๅๅฎ่งๆฐ่" (or similar "menu/help" triggers):
- READ the content of
templates.md in the skill directory.
- DISPLAY the list of available commands to the user exactly as they appear in the file.
- GUIDE the user to select a number or copy the command to execute.
Smart Time Filtering & Reporting (CRITICAL)
If the user requests a specific time window (e.g., "้ๅป X ๅฐๆ") and the results are sparse (< 5 items):
- Prioritize User Window: First, list all items that strictly fall within the user's requested time (Time < X).
- Smart Fill: If the list is short, you MUST include high-value/high-heat items from a wider range (e.g. past 24h) to ensure the report provides at least 5 meaningful insights.
- Annotation: Clearly mark these older items (e.g., "โ ๏ธ 18h ๅ", "๐ฅ 24h ็ฑ้ป") so the user knows they are supplementary.
- High Value: Always prioritize "้ๅคงๆฟ็ญ", "ๅคฎ่กๅๆ
", or "้้ตๆธๆ" items even if they slightly exceed the time window.
Response Guidelines (CRITICAL)
Format & Style:
- Language: ็น้ซไธญๆ (zh-TW).
- Style: Magazine/Newsletter style (e.g., "่ฒก่จ" or "่ฏ็พ่กๆฅๅ ฑ" vibe). Professional, concise, yet engaging.
- Structure:
- ๐ฅ ้ ญๆข็ฆ้ป: Top 3-5 most critical macro stories.
- ๐ฐ ๅคฎ่ก่่ฒจๅนฃๆฟ็ญ: ๅฉ็ใๆตๅๆง็ธ้.
- ๐ ็ถๆฟๆธๆ: GDPใPMIใCPI ็ญๆธๆ่งฃ่ฎ.
- ๐ฑ ๅฏ็่ๅธๅ ด: ไบบๆฐๅนฃใๅตๅธใ่กๅธ็ธ้.
- Item Format:
- Title: MUST be a Markdown Link to the original URL.
- โ
Correct:
### 1. [ๅคฎ่กๅฎฃๅธ้ๆบ 0.5 ๅ็พๅ้ป](https://...)
- โ Incorrect:
### 1. ๅคฎ่กๅฎฃๅธ้ๆบ 0.5 ๅ็พๅ้ป
- Metadata Line: Must include Source, Time/Date, and Heat/Score.
- 1-Liner Summary: A punchy, "so what?" summary.
- Deep Interpretation (Bulleted): 2-3 bullet points explaining why this matters, technical details, or context. (Required for "Deep Scan").
Output Artifact:
- Always save the full report to
reports/ directory with a timestamped filename (e.g., reports/china_macro_YYYYMMDD_HHMM.md).
- Present the full report content to the user in the chat.
- CRITICAL: Report footer MUST include attribution line.
ๆธๆๆบ่ชชๆ
| ไพๆบ |
่ชชๆ |
้ฉ็จๅ ดๆฏ |
| ่ฏ็พ่กๆฅๅ ฑ |
ไธญๅ้ ็ด่ฒก็ถๅช้ซ๏ผๅฎ่ง/ๅธๅ ดๆฐ่ๅณๆๆงๅผท |
ๅคฎ่กๆฟ็ญใๅธๅ ดๅๆ
ใๆธๆ่งฃ่ฎ |
| 36ๆฐช |
็งๆ่ฒก็ถๅช้ซ๏ผๆถต่ๅฎ่ง็ถๆฟๅฟซ่จ |
็ถๆฟๆฟ็ญใ็ขๆฅญๅๆ
|
็ฏไพ่ผธๅบ
# ไปๆฅไธญๅๅฎ่งๆฐ่ๆ่ฆ๏ผ2026-01-20๏ผ
> ๆๆๆ้๏ผ11:30 | ไพๆบ๏ผ่ฏ็พ่กๆฅๅ ฑใ36ๆฐช | ๅ
ฑ 12 ๆข็ธ้ๆฐ่
---
## ๐ฅ ้ ญๆข็ฆ้ป
### 1. [ๅคฎ่กไปๆฅ้ๅฑ 5000 ๅ MLF ๆไฝ๏ผๅฉ็ๆๅนณ](https://wallstreetcn.com/...)
๐ ่ฏ็พ่กๆฅๅ ฑ | ๐ 09:45 | ๐ฅ ้ซ้ๆณจ
ๅคฎ่ก็ถญๆ MLF ๅฉ็ไธ่ฎ๏ผ็ฌฆๅๅธๅ ด้ ๆใ
- **ๆ ธๅฟ่ฆ้ป**๏ผๆฌๆ MLF ๅฐๆ้ 4500 ๅ๏ผๆทจๆๆพ 500 ๅ
- **ๅธๅ ดๅฝฑ้ฟ**๏ผ็ญๆๆตๅๆง็ถญๆๅฏฌ้ฌ๏ผLPR ๅคงๆฆ็ๆๅนณ
- **ๅพ็บ่งๅฏ**๏ผ้ๆณจๆๆซ่ณ้้ข่ไธๆ้ๆบ็ชๅฃ
### 2. [12 ๆ PMI ๅๅ่ณ 50.1๏ผ่ฃฝ้ ๆฅญ้่ฟๆดๅผตๅ้](https://wallstreetcn.com/...)
๐ ่ฏ็พ่กๆฅๅ ฑ | ๐ 10:00 | ๐ฅ ้่ฆๆธๆ
ๅฎๆน่ฃฝ้ ๆฅญ PMI ๅฐๅน
ๅๅ๏ผ็ตๆ้ฃ็บๅ
ฉๅๆๆถ็ธฎใ
- **ๆธๆไบฎ้ป**๏ผๆฐ่จๅฎๆๆธๅๅ 0.3 ๅ็พๅ้ป
- **็ตๆงๅๅ**๏ผๅคงๅไผๆฅญ็ฉฉๅฅ๏ผไธญๅฐไผๆฅญไปๆฟๅฃ
- **ๆฟ็ญๅซ็พฉ**๏ผ็ฉฉๅข้ทๆฟ็ญๆๆๅ้กฏ๏ผไฝๅบ็คๅฐไธ็ฉฉๅบ
---
## ๐ฐ ๅคฎ่ก่่ฒจๅนฃๆฟ็ญ
### 3. [1 ๆ LPR ๅ ฑๅนๅบ็๏ผ1 ๅนดๆ 3.10%ใ5 ๅนดๆ 3.60% ๅๆๅนณ](https://...)
...
---
*ๅ ฑๅ็ฑ list-china-today-macro-news skill ่ชๅ็ๆ*
*๐ Powered by [news-aggregator-skill](https://github.com/anthropics/news-aggregator-skill)*
Attribution
This skill is built upon and extends the architecture of news-aggregator-skill.
- Core fetching patterns derived from
news-aggregator-skill/scripts/fetch_news.py
- Report formatting follows the news-aggregator-skill Response Guidelines
- Smart Time Filtering logic adapted from news-aggregator-skill
๐ Based on news-aggregator-skill by Anthropic
