AgentSkillsCN

document-hub-read

阅读并总结文档中心的当前状态,快速概览系统架构、模块职责、技术栈,以及术语表中的关键概念。

SKILL.md
--- frontmatter
name: document-hub-read
description: Read and summarize the current state of the documentation hub. Provides a quick overview of system architecture, module responsibilities, technology stack, and glossary terms.

Document Hub: Read

View and summarize the current documentation hub state.

Helper Scripts Available:

  • scripts/validate_hub.py - Validates documentation structure

Use this skill to quickly understand a project's documentation without reading all files individually.

What This Skill Does

Reads all documentation hub files and presents a structured summary:

  1. Validates the hub structure
  2. Extracts key information from each file
  3. Presents an organized overview
  4. Identifies any gaps or issues

Decision Tree: When to Use This Skill

code
User wants to understand docs → Does hub exist?
    ├─ No → Suggest /document-hub initialize
    │
    └─ Yes → Read and summarize:
        1. Validate structure first
        2. Read all four core files
        3. Extract key sections
        4. Present formatted summary
        5. Identify any warnings/gaps

Read Workflow

Step 1: Validate Before Reading

Always validate first to catch structural issues:

bash
python scripts/validate_hub.py /path/to/project

If validation returns warnings or errors, include them in the summary.

Step 2: Read Core Files

Read all four core files:

  • systemArchitecture.md
  • keyPairResponsibility.md
  • glossary.md
  • techStack.md

Step 3: Extract Key Information

From each file, extract:

systemArchitecture.md:

  • High-level system description
  • Number of Mermaid diagrams
  • Key architectural components

keyPairResponsibility.md:

  • Project purpose
  • Number of documented modules
  • Module names and responsibilities

glossary.md:

  • Total number of terms
  • Sample key terms (top 5-10)

techStack.md:

  • Core technologies listed
  • Infrastructure components
  • Development tools

Step 4: Present Summary

Format the summary in a clear, scannable structure:

code
Documentation Hub Summary
=========================

Status: ✓ Valid (or ⚠ Issues detected)

## System Architecture
- Description: [1-2 sentence summary]
- Diagrams: 3 Mermaid diagrams
- Key Components: API Gateway, Auth Service, Database

## Module Responsibilities
- Total Modules: 8
- Key Modules:
  • auth - User authentication and authorization
  • payments - Payment processing integration
  • notifications - Email and push notifications

## Technology Stack
- Framework: Next.js 14
- Database: PostgreSQL + Redis
- Infrastructure: Docker, AWS
- [X more technologies]

## Glossary
- Total Terms: 42
- Key Terms: FulfillmentJob, CIP, Ledger, BatchProcessor, ...

## Health Check
- Validation: ✓ Passed
- Warnings: 1 (Complex diagram in systemArchitecture.md)
- Last Updated: [Git last modified date if available]

Example: Complete Read Operation

python
import json
import subprocess
from pathlib import Path

project_path = Path("/path/to/project")
docs_path = project_path / "cline-docs"

# Step 1: Validate
validate = subprocess.run(
    ["python", "scripts/validate_hub.py", str(project_path)],
    capture_output=True, text=True
)
validation = json.loads(validate.stdout)

# Step 2: Check if hub exists
if not docs_path.exists():
    print("Documentation hub not found.")
    print("Run /document-hub initialize to create it.")
    exit(0)

# Step 3: Read files
files = {
    "arch": docs_path / "systemArchitecture.md",
    "resp": docs_path / "keyPairResponsibility.md",
    "glossary": docs_path / "glossary.md",
    "tech": docs_path / "techStack.md"
}

content = {}
for key, file_path in files.items():
    if file_path.exists():
        with open(file_path) as f:
            content[key] = f.read()

# Step 4: Extract and summarize
# [Parse content and extract key information]

# Step 5: Present summary
print("Documentation Hub Summary")
print("=" * 50)
print(f"Status: {'✓ Valid' if validation['valid'] else '⚠ Issues'}")
# [Rest of summary...]

Use Cases

Quick Onboarding

When joining a project:

code
Developer: "What does this project do?"
→ Run /document-hub read
→ Get instant architecture + module overview

Pre-Task Context

Before starting work:

code
Developer: "I need to understand the auth system"
→ Run /document-hub read
→ See auth module responsibility
→ Check glossary for domain terms

Documentation Health Check

Periodic maintenance:

code
→ Run /document-hub read
→ Check for validation warnings
→ If drift detected, run /document-hub analyze

Best Practices

  • Always validate first - Catch structural issues
  • Present concisely - Users want overview, not full content
  • Highlight warnings - Draw attention to validation issues
  • Suggest actions - If issues found, suggest next steps

Common Pitfalls

Don't just dump file contents - summarize them ❌ Don't skip validation - it catches important issues ❌ Don't hide warnings - users need to know about problems

Do provide structured, scannable output ✅ Do include health status upfront ✅ Do suggest follow-up actions if needed

What Comes Next

After reading:

  • If gaps found → Suggest /document-hub analyze for detailed drift report
  • If outdated → Suggest /document-hub update to refresh
  • If invalid → Show specific errors and suggest fixes

Helper Script Reference

validate_hub.py - Check documentation structure

bash
python scripts/validate_hub.py /path/to/project
# Returns: {"valid": bool, "errors": [], "warnings": []}

See scripts/README.md for complete documentation.