skills/agentrs-search/SKILL.md
wangyitong a65adcc2e5 Initial commit: merged, deduplicated, and vetted skill collection
Sources: extracted from two upstream archives (skill-repo, skills-main),
merged with the following policy:

- 15 broken symlinks (pointing to /Users/jameslee/.cc-switch/skills or
  ../../.agents/skills on a foreign machine) discarded
- 3 real name collisions with identical content (ai-pair, ifind-http-api,
  zhipu-websearch) kept as one copy
- Functional overlaps deduped keeping the strongest variant:
  - docx family: kept docx (official, full toolchain) + docx-cn
    (GB/T 9704 Chinese official-document constants),
    dropped docx_writer (no scripts, name collided with docx)
  - humanizer family: kept humanizer-zh (6 zh reference docs),
    dropped humanizer (en, redundant for CN workflow)
- Skills that only ran in a foreign environment removed:
  ablemind-ops, app-publish, hlb-design-system, openclaw-adj-skill,
  claude-driver
- alphapai excluded from this public repo because its SKILL.md hard-coded
  live credentials

Result: 25 skills, 572 files, ~7.5 MB.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-13 14:47:12 +08:00

3.0 KiB
Raw Permalink Blame History

name description
agentrs-search 联网搜索工具箱 —— 通过京东 agentrs 代理网关,用单一 Bearer Token 调用 4 个搜索引擎: tavily(AI搜索)、cloudsway(必应/Bing,可拿网页全文)、SearchPro(搜狗,实时新闻时效最强)、 jdcloud_search(京东云RAG,返回整页正文)。当需要联网检索资讯、新闻、研报、实时信息, 或要抓取网页正文做后续分析时使用。触发词:联网搜索、实时搜索、查资讯、查新闻、 搜一下、tavily、bing、cloudsway、搜狗、京东搜索、web search、抓网页正文。

agentrs-search —— 联网搜索4 引擎)

通过 https://agentrs.jd.com/api/saas/proxy-k/v1/<tool>,单一 Bearer Token 调 4 个搜索引擎。

鉴权与调用约定(先读)

  • 环境变量取密钥,不硬编码export JD_AGENTRS_TOKEN='你的token'
  • 网关只认 JSON body:文档标 GET/POST 都不影响query string 会报"缺少必要输入参数"。
  • HTTP 恒 200业务看 body1001=token 未授权(充值/补参无用);9999 缺少必要输入参数=参数错;9999 用户积分不足=余额不足(重试常能通)。

4 个引擎怎么选

引擎 底层 入参 结果路径 选用场景
tavily Tavily {query} results[] 通用 AI 搜索,默认首选
cloudsway Bing/必应 {q,count,freshness,enableContent,contentType,mainText,sites,blockWebsites} webPages.value[] 中文好;enableContent=true 直接拿网页正文sites/blockWebsites 站点过滤;freshness=Day/Week/Month 时间过滤
searchpro 搜狗 {Query,Cnt,Site,Industry,...} Response.Pages[]元素是字符串化JSON需二次 parse 实时新闻/天气,时效性最强
jdcloud 京东云 RAG {query} data[]page_content 整页正文 + recall_distance 整页全文做分析时

用法

export JD_AGENTRS_TOKEN='...'
python3 scripts/search.py tavily    "人工智能最新趋势"
python3 scripts/search.py cloudsway "中信建投证券" --count 5 --enableContent true --mainText true
python3 scripts/search.py searchpro "今天北京天气"
python3 scripts/search.py jdcloud   "AI agent 发展"

打印规整 JSON遇 1001/9999 在 stderr 给可读提示。

直接 curl

curl -s https://agentrs.jd.com/api/saas/proxy-k/v1/tavily_search \
  -X POST -H "Content-Type: application/json" \
  -H "Authorization: Bearer $JD_AGENTRS_TOKEN" \
  -d '{"query":"搜索关键词"}'

工具名换 cloudsway/SearchPro/jdcloud_search、body 换对应入参即可。

选用范式

  • 通用检索 → tavily;要最新时效 → searchpro;要网页正文做分析 → jdcloud(或 cloudsway --enableContent true);要限定站点 → cloudsway --sites baijiahao.baidu.com

地图/经纬度/POI/路径 等见姊妹 skill agentrs-map;天气见 agentrs-weather。三者共用同一个 JD_AGENTRS_TOKEN。