← 返回 Skills 市场
laigen

A-Share Global Peer

作者 赖根 · GitHub ↗ · v1.0.2 · MIT-0
cross-platform ⚠ suspicious
75
总下载
1
收藏
0
当前安装
3
版本数
在 OpenClaw 中安装
/install a-share-global-peer
功能描述
Find global peer companies for A-share listed companies. Input an A-share company name or code, automatically match overseas listed companies (US, Europe, Ja...
使用说明 (SKILL.md)

A-Share Global Peer Matching Tool

Overview

This skill helps find overseas listed companies that can serve as benchmarks or peers for A-share (Chinese) listed companies. It matches companies based on product/service similarity and provides detailed comparison reports.

Dependencies

Required Tools

Tool Purpose Priority
web_search Search for global leaders, market share data Required
exec Run Python scripts for data fetching Required
read Read reference files Optional

Optional Skills (Enhancement)

Skill Purpose Benefit
ifind-finance-data Query A-share company products/financials Highest accuracy
tushare-data A-share company basic info Structured data
akshare-stock A-share market data Alternative data source

Environment Variables

Variable Required Description
TUSHARE_TOKEN Optional Tushare API token. If not set, falls back to Web Search for company data
BRAVE_API_KEY Optional Brave Search API key. System default available

Note: This skill works without any API keys by using Web Search fallback. API keys enhance data quality but are not mandatory.

Required Files (Bundled)

File Purpose
scripts/get_company_products.py Fetch A-share company product info
references/global_leaders_reference.md Quick lookup table for global leaders

Workflow

Step 1: Get A-Share Company Product Matrix

Data Source Priority: iFinD (if skill available) > Tushare (if token set) > Web Search (always available)

Execute:

python3 scripts/get_company_products.py --company "\x3Ccompany_name_or_code>" --output json

The script automatically:

  1. Checks for TUSHARE_TOKEN environment variable
  2. If available, queries Tushare API
  3. If unavailable, returns recommended Web Search queries

Expected output:

  • Main products/services (sorted by revenue %)
  • Industry classification
  • Revenue geography breakdown

Fallback: If script unavailable or no API token, use web_search with query:

  • "\x3C公司名> 主营业务 营收构成 2024"
  • "\x3C公司名> products revenue breakdown"

Step 2: Search for Global Peer Companies

Search Strategy:

  1. For each major product, search "global leader in [product_name]" or "[product_name] market share"
  2. Prioritize US-listed companies (data easier to obtain), then Europe, Japan, Korea
  3. Use web_search tool, label source as "Web Search"

Selection Criteria:

  • MUST be publicly listed (exclude private companies)
  • Prefer highest market share in same product category
  • Higher product matrix overlap = higher priority

Step 3: Handle Special Cases

Region-Specific Products

When A-share company products have strong regional/Chinese characteristics:

  • Find overseas products with similar regional characteristics for analogy
  • Examples:
    • Baijiu (白酒) → Whisky leader (Diageo)
    • Chinese condiments → Western condiments (McCormick)
  • Mark as "类比对标 (Analogy Benchmark)" and explain logic

Diversified Business

When A-share company operates multiple product categories:

  • First try to find single overseas company with highest product matrix overlap
  • If impossible, benchmark by main product separately
  • Example: Xiaomi (phone + IoT) → Apple (phone + IoT ecosystem), not separate benchmarks

Step 4: Get Market Share Data

Search patterns:

"[company_name] [product] market share"
"[product] global market share ranking"
"[product] industry leaders 2024"

Label all data sources:

  • 数据来源: Web Search (YYYY-MM-DD)
  • 数据来源: iFinD API
  • 数据来源: Tushare API

Step 5: Output Comparison Report (Chinese)

Important: All output reports must be in Chinese for user readability.

Output template:

## A股公司 vs 海外对标

### A股公司:[公司名] ([股票代码])
- **主营产品**: [产品1] (营收占比 XX%), [产品2] (营收占比 XX%)
- **行业**: [行业分类]
- **地域特色**: [如有则说明]

### 海外对标 1:[公司名] ([股票代码] - [交易所])
- **主营产品**: [产品1] (营收占比 XX%), [产品2] (营收占比 XX%)
- **市场地位**: [描述]
- **市场份额数据**:
  - [产品1] 全球份额: XX% (数据来源: ...)
  - [产品2] 全球份额: XX% (数据来源: ...)

### 海外对标 2:[公司名] ([股票代码] - [交易所])
- **主营产品**: ...
- **市场地位**: ...

### 产品对比

| 产品维度 | A股公司 | 海外对标1 | 海外对标2 |
|---------|---------|----------|----------|
| [维度名] | [数值] | [数值] | [数值] |
| 毛利率 | XX% | XX% | XX% |
| 营收规模 | XX亿 | XX亿 | XX亿 |

### 对标逻辑说明
- [选择这家公司对标的原因]
- [地域特色类比依据,如有]

---
*数据获取时间: YYYY-MM-DD*

Examples

Example 1: CATL (宁德时代)

Input: "宁德时代的海外对标公司"

Workflow:

  1. Query CATL main products: EV batteries (~70% revenue)
  2. Search "global leader in EV battery"
  3. Results: LG Energy Solution, Panasonic, Samsung SDI
  4. Get market share data
  5. Output Chinese comparison report

Example 2: Kweichow Moutai (贵州茅台)

Input: "茅台对标哪家海外公司"

Workflow:

  1. Identify Moutai products: Baijiu (strong regional characteristic)
  2. Search "global leader in spirits"
  3. Result: Diageo (whisky leader)
  4. Mark as "类比对标: 白酒 vs 威士忌"
  5. Output analogy comparison report in Chinese

Example 3: Innolight (中际旭创)

Input: "中际旭创的海外对标"

Workflow:

  1. Query: Optical transceivers (95.9% revenue)
  2. Search "global optical transceiver market leader"
  3. Results: Coherent, Lumentum
  4. Special finding: Innolight is already #1 globally!
  5. Output competition benchmark report, note leadership position

Resource Files

scripts/

  • get_company_products.py - Fetch A-share company product info via Tushare or recommend Web Search fallback

references/

  • global_leaders_reference.md - Quick lookup table for global industry leaders

Read reference file to reduce redundant searches:

# Quick lookup example
- EV Battery → LG Energy Solution (373220.KRX), Panasonic (6752.TSE)
- Spirits → Diageo (DEO.NYSE), Brown-Forman (BF.A.NYSE)
- Optical Transceiver → Coherent (COHR.NYSE), Lumentum (LITE.NYSE)

Data Source Labeling Rules

Data Type Label Format
A-share company products 数据来源: iFinD API or 数据来源: Tushare API or 数据来源: Web Search
Overseas company info 数据来源: Web Search (YYYY-MM-DD)
Market share 数据来源: Web Search (YYYY-MM-DD) + search keywords
Estimates/speculation Explicitly label 推测数据,需验证

Important Rules

  1. NEVER fabricate data - All market share figures must come from actual search results
  2. Label uncertainty - Clearly state when data is incomplete
  3. Max 2 peer companies - Avoid information overload, select most representative
  4. Prioritize listed companies - Exclude private companies (e.g., Bosch, Deloitte)
  5. Output in Chinese - User-facing reports must be in Chinese for readability
  6. Check leadership position - Some A-share companies may already be global leaders (e.g., Innolight, CATL)
  7. Graceful fallback - If API tokens unavailable, use Web Search instead of failing

Special Cases Handling

Case 1: A-Share Company is Global Leader

When A-share company already ranks #1 globally:

  • Change framing from "对标学习" to "竞争对标" (competition benchmark)
  • Note that overseas companies are competitors, not learning targets
  • Focus on competitive dynamics, technology divergence, customer overlap

Example: 中际旭创 (Innolight) surpassed Coherent to become #1 optical transceiver supplier in 2024.

Case 2: No Direct Global Peer

When no listed overseas company produces exact same product:

  • Use analogy benchmarking
  • Explain analogy logic clearly
  • Note limitations of analogy comparison

Example: 中国中药 → Western herbal supplements (GNC, privatized, limited options)

Case 3: Multiple Product Categories

When company has diversified business:

  • Try to find single comprehensive peer first
  • If impossible, benchmark main product only
  • Note "多元化经营,仅对标主营业务"

Quality Checklist

Before outputting report, verify:

  • All revenue percentages backed by data sources
  • Market share figures labeled with source
  • Overseas companies confirmed as publicly listed
  • Special cases (regional products, diversified) handled properly
  • Report language is Chinese
  • Data timestamp included
  • If A-share company is global leader, noted in report
安全使用建议
This skill appears to implement its stated purpose, but there are minor inconsistencies you should check before enabling it: 1) The registry metadata shows environment entries as '[object Object]' — ask the publisher or registry maintainer to fix the manifest so the platform won't incorrectly request or expose env variables. 2) Confirm how the web_search tool will run in your environment and whether it relies on BRAVE_API_KEY or other search credentials (the bundled script doesn't use BRAVE_API_KEY, but the skill mentions it). 3) If you plan to supply a TUSHARE_TOKEN, ensure you trust Tushare usage; the token is optional and the skill can work with web search fallback. 4) Review the behaviors of any referenced optional skills (ifind-finance-data, tushare-data, akshare-stock) because those can increase the blast radius by requesting additional permissions or credentials. If you need higher assurance, request a corrected manifest (env vars fixed) and ask the publisher to remove or document any unused environment variables.
功能分析
Type: OpenClaw Skill Name: a-share-global-peer Version: 1.0.2 The skill is a legitimate financial analysis tool designed to match Chinese A-share companies with global peers. It utilizes a Python script (scripts/get_company_products.py) to interface with the Tushare API and provides structured instructions in SKILL.md for conducting web searches and generating comparison reports. No evidence of malicious intent, data exfiltration, or harmful prompt injection was found; the use of environment variables (TUSHARE_TOKEN) and external tools is consistent with the stated functionality.
能力评估
Purpose & Capability
Name/description align with included files: SKILL.md describes web_search + exec to run a Python helper and a reference table; the bundled script fetches company info (Tushare optional) and the reference file lists global leaders. Required tools (web_search, exec) and optional finance skills are proportional to the stated purpose.
Instruction Scope
SKILL.md instructions are scoped to searching and composing comparison reports. The runtime steps instruct using web_search and running scripts; they do not instruct reading unrelated local files or exfiltrating secrets. One caveat: the SKILL.md references iFinD and other skills for higher-quality data—you should confirm what those skills do when invoked (they could request additional permissions/credentials).
Install Mechanism
Instruction-only with one small Python helper file and no install spec. No remote downloads or installers; lowest install risk. The script optionally relies on the tushare Python package but falls back to recommending web search queries.
Credentials
SKILL.md lists two optional env vars (TUSHARE_TOKEN and BRAVE_API_KEY). The bundled script only reads TUSHARE_TOKEN; it does not reference BRAVE_API_KEY. Registry metadata shows required env entries as '[object Object]' which is a serialization/metadata bug and could cause the platform to surface env variables incorrectly. Overall the number and type of env vars are reasonable for the task, but the metadata mismatch and an advertised-but-unused BRAVE_API_KEY are inconsistencies that should be fixed/verified before use.
Persistence & Privilege
always is false; skill is user-invocable and allows normal autonomous invocation. It does not request persistent system-level privileges or attempt to modify other skills' config. No indicators it writes outside its own files.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install a-share-global-peer
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /a-share-global-peer 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.2
Version 1.0.2: No changes, republish to sync
v1.0.1
Fix: Add proper env var declarations in metadata, remove hardcoded paths, add graceful Web Search fallback
v1.0.0
Initial release: Match A-share companies with global peers, output comparison reports with market share data
元数据
Slug a-share-global-peer
版本 1.0.2
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 3
常见问题

A-Share Global Peer 是什么?

Find global peer companies for A-share listed companies. Input an A-share company name or code, automatically match overseas listed companies (US, Europe, Ja... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 75 次。

如何安装 A-Share Global Peer?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install a-share-global-peer」即可一键安装,无需额外配置。

A-Share Global Peer 是免费的吗?

是的,A-Share Global Peer 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

A-Share Global Peer 支持哪些平台?

A-Share Global Peer 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 A-Share Global Peer?

由 赖根(@laigen)开发并维护,当前版本 v1.0.2。

💬 留言讨论