mirror of https://git.openapi.site/https://github.com/desirecore/market.git synced 2026-09-05 18:43:51 +08:00

Files

joy1129-com c83f917901 feat(skills): add 6 due-diligence data-source skills (#111 )

## Summary

Adds six enterprise due-diligence data-source skills as local built-in
skills (SKILL.md + SKILL.zh-CN.md + catalog-metadata sidecar), per
ADR-038.

| Skill | Category | Risk | Source |
|---|---|---|---|
| baidu-poi-search | data | low | Baidu Map Place API (REST) |
| tianyancha-risk | business | low | Tianyancha risk API (REST) |
| ccgp-gov-procurement | business | medium | ccgp.gov.cn (WebBridge) |
| cnipa-patent-search | business | medium | CNIPA pss-system (WebBridge)
|
| creditchina-query | business | high | creditchina.gov.cn (WebBridge +
ddddocr) |
| multi-source-sentiment | research | medium | Douyin Index/Featured +
WebSearch |

## Changes
- 6 x skills/<id>/{SKILL.md, SKILL.zh-CN.md, catalog-metadata.v1.json}
- builtin-skills.json: 34 -> 40
- manifest.json: totalSkills 63 -> 69
- README.md counts + list

## Security
- Credentials replaced with env-var placeholders (BAIDU_MAP_AK,
TIANYANCHA_TOKEN); verified no secrets in diff

## Validation
- validate-i18n.py: 0 errors (remaining warnings pre-existing repo-wide)
- zh/en heading parity verified; source_hash computed per repo algorithm

Co-authored-by: yi-ge <yi-ge@desirecore.net>

2026-09-03 14:05:58 +08:00

7.8 KiB

Raw Blame History

name, description, version, type, risk_level, status, tags, metadata, market

name

description

version

type

risk_level

status

CCGP Government Procurement Search

L0: One-Sentence Summary

Input a company name (short name recommended, e.g. "华为" instead of the full legal name), search ccgp.gov.cn via Kimi WebBridge browser automation, and return a structured list of procurement announcements.

L1: Overview

Data source: China Government Procurement search (search.ccgp.gov.cn)
Auth: none (public search)
Invocation: Kimi WebBridge browser automation (navigate + evaluate)
Prerequisites: Kimi WebBridge daemon running (127.0.0.1:10086) + browser extension connected
Key advantage: a browser session naturally avoids the IP throttling applied to direct curl calls

L2: Procedure

1. Check the WebBridge daemon

curl -s -X POST "http://127.0.0.1:10086/command" \
  -H "Content-Type: application/json" \
  -d '{"action":"list_tabs","session":"ccgp-query"}' --max-time 10

If unreachable, start the daemon:

~/.kimi-webbridge/bin/kimi-webbridge.exe start

2. Navigate to the search results page

Windows approach (write JSON to a temp file to avoid encoding issues):

Write C:\tmp\ccgp-nav.json:

{
  "action": "navigate",
  "args": {
    "url": "https://search.ccgp.gov.cn/bxsearch?searchtype=1&page_index=1&bidSort=0&pinMu=0&bidType=0&dbselect=bidx&kw={company-short-name-URL-encoded}",
    "newTab": true,
    "group_title": "政府采购网查询"
  },
  "session": "ccgp-query"
}

curl.exe -s -X POST "http://127.0.0.1:10086/command" \
  -H "Content-Type: application/json" \
  --data-binary "@C:\tmp\ccgp-nav.json" --max-time 30

Wait 5 seconds before continuing.

3. Extract the announcement list

Write C:\tmp\ccgp-extract.json:

{
  "action": "evaluate",
  "args": {
    "code": "(function(){var txt=document.body.innerText||'';var total=(txt.match(/共找到\\s*(\\d+)\\s*条/)||[])[1]||'?';var items=[];document.querySelectorAll('a').forEach(function(a){var t=(a.innerText||'').replace(/\\s+/g,' ').trim();if(t.includes('{company-keyword}')&&t.length>10){var li=a.closest('li')||a.closest('div');var date='';if(li){var m=(li.innerText||'').match(/\\d{4}[-:]\\d{2}[-:]\\d{2}/);if(m)date=m[0];}items.push({title:t.slice(0,80),href:(a.href||'').slice(0,100),date:date});}});return JSON.stringify({total:total,count:items.length,items:items.slice(0,10)});})()"
  },
  "session": "ccgp-query"
}

curl.exe -s -X POST "http://127.0.0.1:10086/command" \
  -H "Content-Type: application/json" \
  --data-binary "@C:\tmp\ccgp-extract.json" --max-time 40

4. Deep extraction from detail pages (optional)

Navigate to each announcement's detail page and extract structured fields:

{
  "action": "navigate",
  "args": { "url": "{announcement-detail-URL}" },
  "session": "ccgp-query"
}

Wait 5 seconds, then extract:

{
  "action": "evaluate",
  "args": {
    "code": "(function(){var txt=document.body.innerText||'';var lines=txt.split('\\n').map(function(s){return s.trim();}).filter(function(s){return s.length>0;});var fields={};for(var i=0;i<lines.length-1;i++){var l=lines[i];var n=lines[i+1];if(/采购人|采购代理|项目名称|项目编号|采购方式|中标|金额|供应商|地址|联系人|电话|公告日期|品目|更正/.test(l)&&n.length<200){fields[l]=n;}}return JSON.stringify({title:document.title,url:location.href.slice(0,80),fields:fields});})()"
  },
  "session": "ccgp-query"
}

5. Normalized mapping

{
  "query_status": "success",
  "source": "ccgp",
  "company_name": "{company name}",
  "total_found": 2,
  "announcements": [
    {
      "event_type": "招投标公告",
      "event_title": "乌兰察布职业学院物业服务（华为校区）采购更正公告（第一次）",
      "event_date": "2026-08-25",
      "source_url": "http://www.ccgp.gov.cn/cggg/dfgg/gzgg/...",
      "project_number": "WSZC-G-F-260032",
      "purchaser": "乌兰察布职业学院",
      "agency": "乌兰察布市公共资源交易中心",
      "contact": "栗春雨 0474-8305228",
      "correction": "投标截止时间 2026-09-02 → 2026-09-10"
    }
  ],
  "timestamp": "2026-08-31T21:00:00+08:00"
}

6. Pagination (optional)

If total > 10, change page_index=2 in the URL and repeat steps 2-3. Keep ≥5-second intervals.

7. Close the session

{
  "action": "close_session",
  "session": "ccgp-query"
}

Known Limitations

Default time window is about one week: results may be 0 — add start_time and end_time to the URL (e.g. start_time=2026:01:01&end_time=2026:08:31) or click the page's time filter
Title-keyword search: matching happens on announcement titles, not a supplier dimension — full legal names rarely hit; use the short name
Polite pacing: even browser sessions should keep ≥5 s between requests; do not hammer pagination
HTML coupling: redesigns may break selectors — if extraction is empty, fall back to reading document.body.innerText

Troubleshooting

Problem	Cause	Fix
navigate fails	daemon down / extension disconnected	restart via `kimi-webbridge.exe start`
"频繁访问" (frequent-access)	IP throttling	browser sessions usually avoid it; wait 60 s and retry
Empty extraction	0 results / selectors broken	check the keyword; fall back to `evaluate` reading `body.innerText`
Empty detail fields	async content not loaded	retry after 5 s

7.8 KiB Raw Blame History Unescape Escape