doc-orchestrator

DocOrchestrator - Documentation Orchestration Operator

Task Objectives

• This Skill is used for: managing all types of documentation for software development projects throughout their full lifecycle
• Capabilities include: project analysis, document planning, document generation, quality validation, change maintenance, format conversion, intelligent diagramming, version management
• Trigger conditions: when the user needs "create project documentation", "generate requirements documentation", "draw architecture diagrams", "check document consistency", "requirements change analysis", "document version management", etc.

Prerequisites

Dependency Description

This Skill uses the Python 3.8+ standard library, with core dependencies:

• yaml (PyYAML) - dual-track format processing
• json - structured data
• pathlib - path management

Directory Structure

doc-orchestrator/
├── skills/                    # 阶段协调器 + 执行者
├── references/                # 全局参考
├── templates/                 # 统一模板库
└── scripts/                   # 自动化脚本

Operating Steps

Standard Process

1. Analysis Phase

   • Call doc-phase-analysis to analyze project type, scale, and technology stack
   • Output: project-profile.yaml

2. Planning Phase

   • Call doc-phase-planning to plan the document list based on the analysis results
   • Output: doc-plan.yaml

3. Production Phase

   • Call doc-phase-production to generate the document family according to the plan
   • Output: complete document set

4. Quality Phase

   • Call doc-phase-quality to perform consistency and completeness checks
   • Output: quality-report.yaml

5. Maintenance Phase

   • Call doc-phase-maintenance to manage versions, changes, and synchronization
   • Output: updated document set + CHANGELOG.md

Quick Paths

• Small Project: analysis → production → quality (skip planning)
• Existing Project: analysis (scan) → maintenance (sync) → quality
• Diagramming Only: production (diagram-producer)
• Validation Only: quality

Resource Index

Phase Coordinators

Coordinator	Responsibility	Path
doc-phase-analysis	Project/requirements/current-state analysis	View
doc-phase-planning	Document planning/template selection	View
doc-phase-production	Document generation/diagramming	View
doc-phase-quality	Consistency/completeness/scoring	View
doc-phase-maintenance	Version/change/synchronization/export	View

Global References

Reference Document	Purpose	Path
Document Type Catalog	View all supported document types	View
Dual-Track Format Protocol	Agent and human format definitions	View
Intelligent Diagramming Strategy	TXT/Mermaid selection rules	View
Document SOLID Principles	Design principles guide	View
Template Variable System	Template variable definitions	View
Quality Metric Definition	Scoring Rules	View

Notes

• Strictly follow the three-tier architecture (root → coordinator → executor), hierarchy depth ≤ 3
• All documents must include YAML front matter (dual-track format)
• For diagrams, intelligently select TXT/Mermaid by default, adaptively based on complexity
• When maintaining changes, the traceability chain (traceability field) must be updated
• When the quality score is below the threshold, a warning must be marked
• Small projects automatically trim unnecessary documents

Usage Examples

Example 1: Generate documentation from scratch for a new project

Input: "I have an e-commerce project using React + Node.js + PostgreSQL, and I need complete documentation"

Process:

1. analysis: Identified as a full-stack project, medium scale
2. planning: Plan the document list (SRS, SAD, API, DB design, test strategy)
3. production: Generate documents according to the list, intelligently adding diagrams
4. quality: Verify consistency, output quality report
5. maintenance: Establish traceability chain, initialize version

Output: Complete document set + quality-report.yaml

Example 2: Requirement change impact analysis

Input: "Change the original login to support mobile phone number verification code login, analyze the impact"

Process:

1. maintenance/change-analyzer: Trace affected documents (SRS→SAD→API→test)
2. maintenance/doc-syncer: Synchronously update related documents
3. quality: Re-verify consistency
4. maintenance/version-manager: Update version number

Example 3: Quickly generate an architecture diagram

Input: "Draw a microservices architecture diagram"

Process:

1. analysis/project-analyzer: Confirm the architecture type
2. production/diagram-producer: Intelligently select Mermaid to generate the architecture diagram

Supported Document Types

Requirements Category

• SRS (Software Requirements Specification)
• PRD (Product Requirements Document)
• Functional Requirements Document
• Non-Functional Requirements Document

Architecture Category

• SAD (System Architecture Design)
• TDD (Technical Design Document)
• Platform-specific architecture (Web/Mobile/Desktop/Game)
• Data Flow Diagram

Interface and Data Category

• API Interface Document
• Database Design Document
• Communication Protocol Document (WebSocket/gRPC/Protobuf)

Quality and Operations Category

• Test Strategy and Plan
• Deployment and Operations Manual
• Security and Compliance Document

Basic Category

• README
• CHANGELOG
• CONTRIBUTING

Quality Gates

Check Item	Threshold
Document Completeness	≥ 90%
Terminology Consistency	100%
Traceability Completeness	100% have source/destination
Dual-Track Format	100% include YAML front matter
Diagram Coverage	≥ 80% (complex documents)