feishu-doc-to-dev-spec

Convert Feishu Cloud Documents to R&D Requirements Documents

Task Objectives

This Skill is used to:

Read Feishu cloud document content (supports multiple document links)
Fully parse all content in the document: text, tables, images, code blocks, etc.
Based on the development language and storage structure selected by the user
Convert the product requirements document (PRD) into a structured R&D development requirements document

Prerequisites

Feishu Application Configuration

Before use, you need to create a Feishu application and obtain credentials:

Visit Feishu Open Platform
Create a custom enterprise application
Obtain the App ID and App Secret
Configure application permissions:
- docx:document:readonly - Read document content
- drive:drive:readonly - Read cloud space files
- wiki:wiki:readonly - Read knowledge base (if needed)

Environment Variables

export FEISHU_APP_ID=your-app-id
export FEISHU_APP_SECRET=your-app-secret

Operation Steps

Standard Process

Collect document links
- The user provides one or more Feishu document links
- Supported link formats:
  - https://xxx.feishu.cn/docx/xxxxx - New version documents
  - https://xxx.feishu.cn/docs/xxxxx - Old version documents
  - https://xxx.feishu.cn/wiki/xxxxx - Knowledge base documents
Confirm technical stack
- Ask the user to choose a development language (e.g., Python, Java, Go, Node.js, Rust, etc.)
- Ask about the storage structure (e.g., PostgreSQL, MySQL, MongoDB, Redis, etc.)
- Ask about other technical stack preferences (framework, deployment method, etc.)
Read document content
- Call [INVOKE:fetch_feishu_doc] script
- Pass in the document link list
- The script will return the complete document content, including:
  - Text paragraphs
  - Table data (converted to Markdown tables)
  - Images (downloaded locally and paths returned)
  - Code blocks
  - Ordered/unordered lists
Analyze requirement content
- Identify functional requirements and non-functional requirements
- Extract business rules and constraints
- Organize user stories and acceptance criteria
Generate R&D requirements document
- Based on the technical stack selected by the user
- Generate an R&D document containing the following content:
  - Technical architecture design
  - Data model design (table structure)
  - API interface design
  - Core function implementation plan
  - Technical risk assessment
Output results
- Generate an R&D requirements document in Markdown format
- Save to ./output/ directory

Resource Index

Scripts

Feishu document reading：scripts/fetch_feishu_doc.py

Purpose: Read the complete content of a Feishu cloud document

Input parameters (JSON format):

{
  "doc_urls": ["https://xxx.feishu.cn/docx/xxxxx"],
  "app_id": "飞书应用ID（可选，默认从环境变量读取）",
  "app_secret": "飞书应用密钥（可选，默认从环境变量读取）"
}

Output: JSON structure of the document content

Reference documents

R&D document template：references/dev-spec-template.md
- When to read: When generating a development requirements document
- Contains the standard development document structure and examples
Technology Selection Guide：references/tech-stack-guide.md
- When to read: When helping users choose a technology stack
- Contains the characteristics and applicable scenarios of common technology stacks

Notes

Document Permissions

Ensure the Feishu app has permission to access the target document
If the document is private, the app needs to be added as a document collaborator

Image Handling

Images in the document will be downloaded to the ./output/images/ directory
Image links will be replaced with local relative paths

Multi-Document Merging

When the user provides multiple document links, the content will be read and merged in order
The content of each document will be separated by a divider line

Output Format

The development requirements document uses Markdown format
Tables use Markdown table syntax
Code blocks use syntax highlighting markers for the corresponding language

Usage Examples

Example 1: Single Document Conversion

User Input：

请帮我把这个产品需求文档转换为研发需求文档：
https://example.feishu.cn/docx/abc123

技术栈：
- 语言：Python
- 框架：FastAPI
- 数据库：PostgreSQL
- 缓存：Redis

Execution Flow：

Call [INVOKE:fetch_feishu_doc] to read the document
Analyze the requirements content
Generate the development document based on the Python + FastAPI + PostgreSQL technology stack
Output to ./output/dev-spec.md

Example 2: Merging Multiple Documents