agentsclimarketplace

Hlzd b2b research

Skill Alexxiang2008/hlzd-b2b-export-skills/skills/hlzd-b2b-research

Agent Skills for B2B industrial exporters — market research (HS code, UN Comtrade, Google Trends, tender), buyer finding (Alibaba + Volza customs), inquiry qualification, voice-contract HTML reports. By HLZD (海联智达).

Install
npx -y skills add Alexxiang2008/hlzd-b2b-export-skills --skill hlzd-b2b-research

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • 22 days oldThe repository was created 22 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

B2B 工业品海外市场调研全链路自动化 —— HS 编码查询(hsbianma.com)+ UN Comtrade 市场规模 + Google Trends 需求热度 + DDGS 买家线索 + 招标信息追踪,输出结构化 JSON 与 Markdown 报告。Use when 用户说'B2B 调研'、'工业品市场分析'、'查 HS 编码'、'查采购商'、'找海外买家'、'市场容量'、'招标追踪'、'B2B market research'、'industrial export research'、'find buyers overseas'、'project tender tracking'。

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

10.8 KB, ~3.2k tokens by cl100k_base, as published. Nobody here has run it

HLZD B2B 工业品海外调研

把"卖什么?卖到哪?卖给谁?"在 30 秒内跑出 4 段数据:HS 编码 → 进口规模 → 需求热度 → 真实买家。


When to use

调用本 Skill 当用户:

  • 要把某个工业品卖到海外,需要先判断市场规模 + 中国出口位置
  • 手里有几个目标市场(UAE / 沙特 / 尼日利亚 / 巴西…),需要并行调研
  • 需要 HS 编码 + 监管条件 + 退税率
  • 需要从 Google / LinkedIn 找真实买家(非 Alibaba / Made-in-China 平台供应商)
  • 需要追踪政府 / World Bank / ADB 项目招标
  • 需要把多次调研结果汇总成一份可发给客户的报告

不要调用本 Skill 当

  • 用户只要单一数据点(如"今天美元汇率")—— 直接用单子脚本
  • 用户只要 LLM 文本生成(不要当成 prompt 模板)
  • 消费品出海(应改用 hlzd-market-report

How this skill is invoked

/hlzd-b2b-research "<product>" [markets...]

单独子能力

任务直接调用
HS 编码py scripts/hs_lookup.py --keyword "石油套管"
Comtrade 贸易数据py scripts/trade_data.py --hs 730429 --reporter sa --period 2023
Google Trendspy scripts/keyword_trends.py --kw "OCTG pipe" "API 5CT" --geo US --delay 12
买家 + 招标py scripts/buyer_search.py --query "container house mining" --market "Australia Africa" --tenders --max 30

完整 pipeline(一键出报告)

py scripts/run_research.py \
    --product "OCTG casing" \
    --markets "UAE Saudi" \
    --hs-candidate 730429 \
    --output-md report.md \
    --output-json report.json \
    --pretty

输出示例:

{
  "product": "OCTG casing",
  "markets": ["UAE", "Saudi"],
  "hs_code_used": "730429",
  "hs_code_candidates": [...],
  "trade_data": { "UAE": { ... }, "Saudi": { ... } },
  "trends": { "UAE": { ... }, "Saudi": { ... } },
  "buyers": { "UAE": { "total_buyers": 18, "buyers": [...] },
              "Saudi": { "total_buyers": 12, "buyers": [...] } },
  "warnings": []
}

The 4-step pipeline

                  ┌──────────────────┐
   product name ─→│  Step 1: HS 编码  │ → 6位 HS Code + 监管条件 + 退税率
                  └────────┬─────────┘
                           ↓
                ┌──────────────────────────┐
                │   per-market (循环)       │
                │                           │
                │  Step 2: UN Comtrade      │ → 进口总额 + 中国份额排名
                │  Step 3: Google Trends    │ → 平均热度 + 相关上升词
                │  Step 4: DDGS 买家 + 招标  │ → 真实买家 URL + 类型
                └──────────────────────────┘
                           ↓
              ┌─────────────────────────┐
              │ 报告 (JSON + Markdown)  │ → data-driven 决策
              └─────────────────────────┘

Step 1 — HS 编码查询(hs_lookup.py)

数据源:hsbianma.com(中文品名 → 完整 10 位 HS 编码 + 监管条件 + 退税率 + 状态)

py scripts/hs_lookup.py --keyword "石油套管" --pages 2

输出字段:

[{
  "hs_code": "7304.29.10",
  "name": "石油套管",
  "unit": "千克",
  "rebate": "0",
  "regulation": "4,x",
  "expired": false
}]

UN Comtrade API 用 HS 编码前 6 位(如 7304.29.10730429)。

Step 2 — UN Comtrade 市场规模(trade_data.py)

数据源:UN Comtrade API(comtradeapi.un.org),免费,无需 Key

py scripts/trade_data.py --hs 730429 --reporter sa --period 2023 --output data.json

常用国家简写:us / ae / sa / ng / za / au / br / mx / cn / world

输出字段(JSON):

{
  "hs_code": "730429",
  "reporter": "sa",
  "total_import_value_usd": 1234567890.0,
  "data_source": "UN Comtrade",
  "countries": [
    { "country": "China", "import_value_usd": 5e8, "share_pct": 41.7 },
    { "country": "Japan", "import_value_usd": 3e8, "share_pct": 25.0 }
  ]
}

关键解读:

  • 进口总额 > 1 亿美元 → 大市场
  • 中国份额 > 30% → 进入有空间但竞争激烈
  • 中国份额 < 10% → 进入机会大但需要认证突破

Step 3 — Google Trends 需求热度(keyword_trends.py)

数据源:Google Trends(pytrends),免费

py scripts/keyword_trends.py --kw "OCTG pipe" "API 5CT casing" --geo US --delay 12

⚠️ 429 限流保护:每个关键词查询间隔 ≥12 秒,触发 429 后等待 90 秒重试 3 次。

输出字段:

{
  "keyword": "OCTG pipe",
  "interest_over_time": { "<date>": 65, ... },
  "related_queries": [ { "query": "API 5CT", "value": "+250%" } ],
  "interest_by_region": { "Saudi Arabia": 100, "UAE": 85 }
}

Step 4 — DDGS 买家 + 招标信息(buyer_search.py)

数据源:DuckDuckGo(ddgs 包),免费无 Key

py scripts/buyer_search.py \
    --query '"OCTG casing"' --market "Saudi Arabia UAE Nigeria" \
    --max 30 --tenders --delay 4

买家类型分类:

类型关键词价值
矿业/油田营地mining / oilfield / camp主力买家,批量采购
EPC/总包EPC / contractor / turnkey项目驱动,决策链长
政府/NGOgovernment / World Bank / ADB大单,需合规认证
贸易商importer / trading house中间渠道,快速起量
酒店/文旅hotel / resort / glamping需求稳定
石油设备商OCTG / API 5CT / line pipe专用品类

内置过滤规则(自动跳过供应商页面):

  • ❌ alibaba / made-in-china / globalsources / ec21
  • ❌ "shipping container"(集装箱房屋类常误匹配)
  • ❌ "door casing"(casing 类常误匹配建筑装饰)

Output schema

完整 pipeline 输出(run_research.py):

{
  "$schema": "hlzd/b2b-research/v1",
  "product": "<原始产品名>",
  "markets": ["<market1>", "<market2>", ...],
  "hs_code_used": "<最终使用的 6 位 HS>",
  "hs_code_candidates": [ {...}, ... ],
  "trade_data": {
    "<market>": {
      "hs_code": "...",
      "reporter": "...",
      "total_import_value_usd": 0.0,
      "data_source": "UN Comtrade",
      "countries": [ { "country": "...", "import_value_usd": 0.0, "share_pct": 0.0 }, ... ]
    }
  },
  "trends": {
    "<market>": {
      "keyword": "...",
      "geo": "...",
      "interest_over_time": null or { ... },
      "related_queries": [ ... ],
      "interest_by_region": null or { ... },
      "error": null or "<error msg>"
    }
  },
  "buyers": {
    "<market>": {
      "total_buyers": 0,
      "total_tenders": 0,
      "buyers": [ { "title": "...", "url": "...", "snippet": "...", "buyer_type": "..." } ],
      "tenders": [ ... ]
    }
  },
  "warnings": [ "<step>:<market>: <reason>", ... ]
}

JSON 输出到 report.json,Markdown 报告到 report.md


Sample outputs

assets/sample-outputs/(真实历史调研,脱敏)。


Red-flag handling

风险处置
UN Comtrade 0 记录检查 HS 编码 / 年份;尝试 reporter=world
Pytrends 429 限流自动等待 90 秒重试 3 次;继续运行其它步骤
pytrends 数据 partial正常(Google 对低流量词标记 partial)
DDGS 返回 0 结果换引号短语关键词;查 ddgs 包版本
hsbianma.com 访问失败改用 volza.com / 中国海关编码中心
卖家数据真实性必查 buyer 官网 + LinkedIn 二次复核

Related Skills

              ┌────────────────────────────┐
              │  hlzd-buyer-finder         │  (W2-3 升级中)
              │  基于海关数据找真实进口商    │
              └────────────────────────────┘
                          │ 衔接
                          ▼
              ┌────────────────────────────┐
              │  ★ hlzd-b2b-research ★    │  ← 本 Skill
              │  4 步 pipeline 出报告        │
              └────────────────────────────┘
                          │
        ┌─────────────────┼─────────────────┐
        ▼                 ▼                 ▼
  hlzd-cold-outreach  hlzd-inquiry-     hlzd-market-report
  (用买家名单写邮件)    qualify           (HTML 报告渲染)
                      (对进来的询盘评分)

前置:无(首步) 后继hlzd-buyer-finder / hlzd-cold-outreach / hlzd-inquiry-qualify / hlzd-market-report


Boundaries & Limitations

  • 本 Skill 不是合规判定:制裁国家粗筛仅在 hlzd-inquiry-qualify 末段
  • HS 编码建议:自动爬取仅为初判,争议商品请人工复核中国海关税则 / WCO HS 委员会
  • Comtrade 数据滞后:UN Comtrade 数据滞后约 1.5 年(最新通常为 T-1 年)
  • DDGS 容量:免费版每天约 10 次搜索;多市场轮询要做限额控制
  • 未启用 LLM:所有抽取走确定性规则;LLM 仅用于最终报告文案的润色

Versioning

版本说明
0.1.0首版:4 脚本 + run_research orchestrator + lib.py(共享基础设施)

升级轨迹见仓库根 VERSIONS.md


Crafted for B2B industrial exporters — HLZD Cross-Border AI Platform · 2026

Gives 0 of the 12 instructions most research analysis skills give in ~3.2k tokens

Counted across 1,063 of the 1,754 authors here whose files we hold, read 2026-08-06

  • generate a markdown reportin 32 of 1063, across 17 files
  • cite each claim's sourcein 31 of 1063, across 14 files
  • define the ideal customer profilein 20 of 1063, across 2 files
  • search for companies matching the criteriain 20 of 1063, across 2 files
  • assign a fit score from one to tenin 20 of 1063, across 2 files
  • format results in a scannable markdown templatein 20 of 1063, across 2 files
  • analyze the codebase to understand the productin 19 of 1063, across 1 file
  • ask clarifying questions about the value propositionin 19 of 1063, across 1 file
  • look for signals of immediate needin 19 of 1063, across 1 file
  • identify the target decision maker rolein 19 of 1063, across 1 file
  • suggest a personalized contact strategyin 19 of 1063, across 1 file
  • provide conversation starters for outreachin 19 of 1063, across 1 file

Said here and by no other author read

  • query HS codes from hsbianma.com
  • fetch import scale and China share from UN Comtrade
  • fetch demand heat from Google Trends
  • search for real buyers and tenders via DDGS
  • wait 12 seconds between keyword queries
  • filter out supplier platforms from buyer search

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.