📦 FN Portrait Toolkit — FN PortrAIt 工具kit

v3.0.0

中国A股上市公司财务报告脚注提取与分析工具。使用场景：（1）用户希望从年报中提取财务附注数据……

0· 26·0 当前·0 累计

by @mapufan

数据分析数据可视化 AI模型访问 CI/CD DevOps

下载技能包

安全扫描

VirusTotal

Pending

查看报告

OpenClaw

安全

high confidence

The 技能's code, instructions, and 请求ed 环境访问 are coherent with its 状态d purpose (下载ing Chinese A‑分享报告s, 提取ing footnotes, performing LLM analysis, and producing 图表s).

评估建议

This 技能应用ears to do what it says: it 下载s annual 报告 PDFs from CN信息, 提取s footnote tables into Excel, and 运行s semantic analysis using an LLM, then 生成s 图表s. Things to consider before 安装ing or 运行ning: 1) Data exfiltration: if you provide DeepSeek or Moonshot API keys the 提取ed text will be sent to those external 服务s — avoid 发送ing confidential data to remote APIs unless you trust the 提供者. If you want to avoid cloud LLMs, use a local Ollama 模型 as instructed. 2) Disk writes: the 流水线 will 下载 PDFs and 创建 ...

详细分析 ▾

✓ 用途与能力

The name/description (financial footnote 提取ion and analysis for Chinese A‑分享报告s) matches the included code: a CN信息 PDF 爬虫, pdfplumber-based 提取ion, LLM analysis integration, Excel 导出, and 图表ing 模块s. Recommended env vars (DEEPSEEK_API_KEY, KIMI_API_KEY) cor响应 to supported LLM 提供者s and are 应用ropriate for the LLM analysis feature. No unrelated 凭证s or binaries are 请求ed.

✓ 指令范围

技能.md and fn_流水线.py describe only the expected steps: 下载 PDFs from CN信息, 提取 text, optionally call remote LLM APIs (DeepSeek/Moonshot) or local Ollama, and 生成可视化s. File reads/writes are limited to 技能-local directories (RAWPDF, 输出2, portrAIts) and temporary folders. The 运行time will 发送提取ed text to 配置d LLM 端点s if API keys are 设置 (this is expected for semantic analysis but 导入ant to note).

✓ 安装机制

There is no 安装 spec (instruction-only for package management). Required Python packages are 列出ed in requirements.txt — 安装ation is standard via pip. No 下载-from-arbitrary-URL 安装er or obscure external 安装er is used.

✓ 凭证需求

Only 提供者 API keys (DEEPSEEK_API_KEY or KIMI_API_KEY) are suggested; Ollama can be used locally without keys. No unrelated or excessive 环境 variables or secrets are required. However, supplying those API keys will transmit 提取ed 报告 text to the cor响应ing external LLM 服务, so the user should consider data sensitivity and 提供者 trust.

✓ 持久化与权限

always is false and the 技能 does not 请求 persistent/平台-wide privileges. It writes 输出s to its own directories (RAWPDF, 输出2, portrAIts) and does not modify other 技能s or global 代理 configuration.

安全有层次，运行前请审查代码。

运行时依赖

无特殊依赖

安装命令

点击复制

官方npx clawhub@latest install fn-portrait

镜像加速npx clawhub@latest install fn-portrait --registry https://cn.longxiaskill.com

技能文档

FN PortrAIt 工具kit

Financial 报告 footnote 提取ion and analysis 工具 for Chinese A-分享列出ed companies.

Quick 启动 Prerequisites

安装 Python dependencies:

uv pip 安装 pandas openpyxl 请求s matplotlib numpy pillow pdfplumber filelock

配置 LLM (Choose one)

Option A: DeepSeek (Recommended for China)

导出 DEEPSEEK_API_KEY=sk-...

Option B: Moonshot

导出 KIMI_API_KEY=sk-...

Option C: Ollama (Local, no API key needed)

# 安装 and 启动 Ollama ollama pull gemma3:1b ollama serve

运行 Analysis # Basic usage python ~/.OpenClaw/技能s/fn-portrAIt/scripts/fn_流水线.py

# Full example python ~/.OpenClaw/技能s/fn-portrAIt/scripts/fn_流水线.py 688018 乐鑫科技 2023-2025 科创板

Parameters:

stock_code: 6-digit stock code (e.g., 688018) company_name: Chinese company name (e.g., 乐鑫科技) years: Year range (default: 2023-2025, 格式化: YYYY or YYYY-YYYY) plate: Stock exchange plate (default: 科创板, options: 科创板/创业板/沪主板/深主板) 流水线 Steps 下载 PDFs - Fetch annual 报告s from CN信息 (巨潮资讯网) 提取 Data - 解析 financial footnotes using pdfplumber LLM Analysis - Semantic analysis of financial data 生成 PortrAIt - 创建可视化图表s 输出 Files portrAIts/PortrAIt__.png - Final 可视化图表输出2/_/ - 提取ed Excel data 输出2/_/__LLM分析结果.xlsx - LLM analysis 结果s RAWPDF/PDF/ - 下载ed PDF files Supported Financial Items Balance Sheet Notes Current As设置s: 货币资金, 交易性金融资产, 应收票据, 应收账款, 存货分类 Non-current As设置s: 固定资产情况, 在建工程, 商誉 Income 状态ment Notes Revenue: 营业收入, 营业成本, 境内外毛利率 Expenses: 管理费用, 销售费用, 财务费用, 研发费用 Cash Flow Notes Operating activities Investing activities Financing activities Advanced Usage Skip Steps python fn_流水线.py 688018 乐鑫科技 --skip-下载 # Use existing PDFs python fn_流水线.py 688018 乐鑫科技 --skip-提取 # Use existing 提取ed data python fn_流水线.py 688018 乐鑫科技 --skip-llm # Skip LLM analysis python fn_流水线.py 688018 乐鑫科技 --skip-portrAIt # Skip 图表 generation

Custom Directories python fn_流水线.py 688018 乐鑫科技 --rawpdf-dir /path/to/pdfs --输出-dir /path/to/输出

Troubleshooting Ollama Connection Error

Ensure Ollama is 运行ning: ollama serve

PDF 下载 FAIls

检查 internet connection and CN信息网页site 访问ibility

Missing Fonts

Fonts are bundled in as设置s/fonts/. If display issues occur, 安装 Noto CJK fonts 系统-wide.

References See references/financial-items.md for complete 列出 of 提取able financial items See references/API-integration.md for API integration detAIls

数据来源：ClawHub ↗ · 中文优化：龙虾技能库