agentsclimarketplace

Hlzd market intel

Skill Alexxiang2008/hlzd-b2b-export-skills/skills/hlzd-market-intel

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-market-intel

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

  • 23 days oldThe repository was created 23 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 跨境贸易市场情报 v0.2 — 抓取 Reddit/X/YouTube/TikTok/Bluesky/HackerNews/GitHub/arXiv 等 14+ 平台 30 天真实用户声音,输出 Top 5 痛点 + 用户原话 + 跨境 actionable 的自包含 HTML 报告(融合 4-agent pipeline + Pain Profile 行业预设 + Voice Contract 7 LAWS)。受众:B2B 跨境贸易老板 / 业务团队 / 选品经理。Use when 用户要做海外产品调研 / 选品 / 痛点分析 / 竞品监控 / 客户声音采集。

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

5.7 KB, ~1.6k tokens by cl100k_base, as published. Nobody here has run it

HLZD Market Intel v0.2.0

B2B 跨境贸易市场情报 — 抓 14+ 平台 30 天真实用户声音,输出 Top 5 痛点 + 用户原话 + 跨境 actionable 自包含 HTML。

v0.2 升级亮点 (vs v0.1)

  • 适配仓库 AGENTS.md frontmatter 规范 (name/description/license/metadata.triggered_by)
  • SKILL.md 精简 从 2264 行到 ~150 行;详细 reference 移到 references/SKILL_FULL.md
  • 可选依赖 降到最小:纯 stdlib (json, urllib, csv) 可跑;可选 API (Reddit/X/YouTube/TikTok) 需 key
  • Voice Contract 7 LAWS 兼容 (与 hlzd-market-report 共享)

工作流 (10 步)

  1. Step 0: First-Run Setup Wizard — 配置 API keys (SCRAPECREATORS / OPENAI / XAI / BRAVE 等)
  2. Step 0.45: Query Quality Pre-Flight — 检测 keyword-trap 主题 (健康/政治) → 提前降级
  3. Step 0.5: Pre-Flight Resolution — 解析 handle / repo / community
  4. Step 0.55: Pre-Research Intelligence — 解析社区 + handles (Reddit/HN/GitHub/Bluesky 等)
  5. Step 0.75: Query Plan — LLM 自己当 planner,决定抓哪些 platform + handles
  6. Research Execution — 调 scripts/briefing.py → 14 平台爬虫
  7. Save Raw + WebSearch Augment — 落盘原始 + 补充 WebSearch 结果
  8. Cluster & Fuse — scripts/lib/cluster.py + fusion.py 去重 + 合并 cluster
  9. Entity Extract — scripts/lib/entity_extract.py → 抽 company / product / pain point
  10. HTML Report — scripts/lib/html_render.py → 自包含 HTML 输出

快速使用

CLI

# 默认 30 天窗口 + 14 平台
py scripts/briefing.py "海外安防摄像头 2026 痛点"

# 指定 pain profile + 平台
py scripts/briefing.py "exosomes" --pain-profile presets/pain-profiles/biomedical.yaml

Python API

from scripts.briefing import run_research

result = run_research(
    topic="智能门锁 跨境选品",
    pain_profile="presets/pain-profiles/security.yaml",
    output_html="report.html",
)

4-Agent Pipeline (--agent flag)

Plan Agent → Research Agent → Cluster Agent → Report Agent

详细 agent 协议见 references/SKILL_FULL.md §Agent Mode。

输出 schema

{
  "topic": "...",
  "window_days": 30,
  "platforms_scanned": ["reddit", "hackernews", "youtube", "..."],
  "signals": [
    {
      "platform": "reddit",
      "url": "...",
      "title": "...",
      "snippet": "...",
      "pain_signal": "...",
      "engagement_score": 0.85
    }
  ],
  "clusters": [
    {
      "name": "Top Pain Point",
      "occurrences": 47,
      "user_quotes": ["..."],
      "actionable": "..."
    }
  ],
  "html_report": "report.html"
}

配置文件 (4-Agent pipeline)

config/company.yaml + locale/{en_US,zh_CN}.json + 4 规则 PDF (UCP / ICC / ISP98 / URDG)。

依赖

必装 (零外部依赖):

  • Python 3.10+
  • 库: requests, urllib3 (标准库)

可选 (提升数据来源):

  • SCRAPECREATORS_API_KEY — TikTok / YouTube / Instagram
  • OPENAI_API_KEY / XAI_API_KEY / OPENROUTER_API_KEY — LLM 提炼痛点
  • BRAVE_API_KEY — WebSearch
  • BSKY_HANDLE + BSKY_APP_PASSWORD — Bluesky
  • TRUTHSOCIAL_TOKEN — Truth Social

详见 references/SKILL_FULL.md §Configuration。

测试

py -m pytest tests/ -q          # 全 50+ tests 跑 (含 hermes / adversarial_v3)

注意: 部分测试需要外部 API (Reddit, Twitter 等)。offline 时只跑纯函数子集:

py -m pytest tests/test_categories.py tests/test_dedupe.py -q

Related Skills

  • hlzd-market-report (姊妹) — 9 节 HTML 自包含报告 (本 skill 输出是其上游)
  • hlzd-pipeline-viz (下游) — 跨 skill trace 累计 + dashboard
  • hlzd-b2b-research (姊妹) — HS 编码 + UN Comtrade 市场数据
  • hlzd-customer-profile (姊妹) — 客户 360° 画像

详细文档

  • references/SKILL_FULL.md — 完整 2000+ 行工作流 (Step 0-7 + 7 LAWS + 4 Agent Protocol + 14 平台 backends)
  • references/methodology.md — 5-dim 信号评分 (与 hlzd-market-report 共享)
  • references/save-html-brief.md — 输出 HTML 模板规范
  • references/hlzd-extensions.md — HLZD 私有扩展 (Pain Profile 行业预设)
  • presets/pain-profiles/ — 6 行业预设 (security / biomedical / SaaS / hardware / food / fashion)
  • presets/openai.yaml — OpenAI LLM 提炼 prompt 配置

Versioning

版本说明
0.1.0原始 (mvanhorn/last30days fork)
0.2.0适配仓库规范: frontmatter + SKILL.md 精简 + Voice Contract LAWS

What ships with it: 263 files

17222.2 KB alongside SKILL.md, 248 of them executable

agents/

scripts/

223 more files not listed here. See all 263 in the repository.

Gives 0 of the 12 instructions most social media skills give in ~1.6k tokens

Counted across 489 of the 492 authors here whose files we hold, read 2026-08-07

  • Adapt formats and tone to each platformin 26 of 489, across 14 files
  • Build content around three to five pillarsin 25 of 489, across 13 files
  • Read product marketing context before asking questionsin 23 of 489, across 13 files
  • Respond to all comments on your postsin 21 of 489, across 9 files
  • Use the output flag to specify an output directoryin 14 of 489, across 4 files
  • Generate output logo images with white backgroundin 13 of 489, across 4 files
  • Fix failing generation scripts directlyin 13 of 489, across 4 files
  • Ask user about HTML preview after logo generationin 12 of 489, across 3 files
  • Run the download script with a URLin 12 of 489, across 3 files
  • Implement exponential backoff for 429 responsesin 12 of 489, across 3 files
  • Write the hook firstin 12 of 489, across 7 files
  • Include a single clear call to actionin 12 of 489, across 9 files

Said here and by no other author read

  • Run the first-run setup wizard to configure API keys
  • Run the query quality pre-flight check
  • Resolve specified handles, repositories, or communities
  • Execute pre-research intelligence on target communities
  • Generate a platform and handle query plan
  • Run the research briefing script

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.