canghe-tianyancha
Generates Tianyancha-style company and industry insight dashboards from researched enterprise data. Use when user asks for 天眼查, company lookup, enterprise analysis, hiring analysis, legal risk, financing, shareholders, intellectual property, investments, or industry company comparison.
1How It Works & Mechanism
After entering a company name or industry name, the agent first uses Kimi Work's built-in kimi_search_v2 to retrieve public Tianyancha data in parallel, followed by structured extraction.
天眼查企业洞察看板 Skill —— 输入企业名称，自动生成包含企业概况、在招岗位分析、数据分析、司法风险、融资历史、股东结构、知识产权、对外投资、主要人员的单文件 HTML 看板。
Retrieval is a host capability rather than an implementation of this repo: SKILL.md explicitly states that kimi_search_v2 and PythonRun are built-in Kimi Work tools.
> **导出说明**：本 Skill 已从原 Kimi 环境导出，适配其他 Kimi Work 用户使用。`kimi_search_v2` 和 `PythonRun` 均为 Kimi Work 内置工具，无需修改。
Single-company mode retrieves data in parallel across 6 query types (active recruitment / judicial risk / financing & shareholders / business registration & executives / intellectual property / outbound investments).
- `{企业名} 天眼查 在招岗位 招聘 岗位分析`
Industry mode first retrieves a list of leading enterprises, supplements complete data for each company, and finally batch-generates company dashboards + an industry dashboard.
2. 使用 `kimi_search_v2` 搜索该行业头部企业名单：
Dashboard generation is done with local Python: invoking functions like generate_dashboard() in scripts/generate_dashboard.py to output a single-file HTML.
sys.path.insert(0, '${SKILL_DIR}/scripts')
The output is a single-file HTML with zero backend dependencies, rendering charts using Tailwind CSS + ECharts, viewable directly in a browser.
- 单文件 HTML，零后端依赖，浏览器直接打开
The dashboard has 9 fixed tabs (Company Overview / Active Hiring Analysis / Data Analytics / Judicial Risks / Financing History / Shareholder Structure / Intellectual Property / Outbound Investments / Key Personnel).
看板包含 9 个 Tab，默认显示
Active hiring analysis only fetches the most recent 90 days (one quarter) and performs 7 chart analyses.
**核心规则：只显示最近一个季度（90天）的岗位数据分析。**
The data analysis module algorithmically generates company health scores, hiring trend forecasts, job-seeker friendliness, and comprehensive recommendations.
综合评分（满分100分），从四个维度评估：
The preference file EXTEND.md controls the output directory, analysis window, sources, language, and other settings.
**EXTEND.md Supports**: default output directory, recent job analysis window, industry company count, preferred data sources, report language, and whether to include dashboard disclaimers.
2Core Capabilities
3External Dependencies
| Kind | Dependency / Endpoint |
|---|---|
| api | kimi_search_v2（Kimi Work 内置检索工具，用于获取天眼查公开数据） |
| cli | PythonRun（Kimi Work 内置 Python 执行器，用于运行 generate_dashboard.py） |
4Recommended Governance & Evidence
- scripts/: Execute local script or command: PythonRun executes gener from skills/canghe-tianyancha/scripts/generate_dashboard.py
- 网络端点: Declare external network access: The skill's own script has zero network calls (generate_dashboard.py depends only on the Python standard library, risk scan reported 'Network requests: No matches')
- 本地工作目录: File read/write and artifact output: 企业洞察看板_{企业名}.html (Single company)
5Second-Pass Verification
- [ok] 脚本无网络、无凭证、仅 stdlib — 风险扫描『網路請求：無命中』；README 明写『generate_dashboard.py 只依赖 Python 标准库』；generate_dashboard.py 在 §2 被标记 generated 未展开。
- [ok] 数据获取与执行是宿主工具 — SKILL.md『导出说明』明说 kimi_search_v2 与 PythonRun 为 Kimi Work 内置；按批4 细则不计入档位。
- [ok] 产物为单文件 HTML — SKILL.md 输出节与 README 均写单文件、零后端依赖、Tailwind CSS + ECharts。
- [ok] 看板结构（9 Tab / 7 图 / 90 天） — SKILL.md 逐条列出 9 个 Tab、在招岗位分析 7 个 ECharts 图与『只显示最近一个季度（90天）』核心规则。
- [ok] 算法模块为估算并附免责 — 健康度满分 100 四维度、线性回归预测 4 周、求职者友好度满分 100；SKILL.md/README 均注明仅供参考、不构成建议。
- [ok] 引用字符安全 — 含全角引号的原文（如本季度切换说明）未纳入 evidence.quote，避免逐字复核时的引号歧义。
- [ok] 档位复核 — 自身脚本仅本地生成 HTML → blue；宿主工具/检索外发不计档，已写入 risks。
- [ok] 元数据与 commit — commit=dd0bf355955b4c82b764740b4183c86a72ba0e0c，license=MIT（仓库根），installs=null。
6Verdict & Scenarios
4947418567e8b52f…