study/CLAUDE.md

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Project Overview

This is 学习助手 Pro (Learning Assistant Pro), an enterprise-grade AI-driven development system built with comprehensive quality assurance, governance, and context management. The system specializes in providing Claude AI with a complete development methodology, tools, and workflows for high-quality software delivery.

Core Framework Components

• Claude Bluebook v1.2: Enterprise-grade AI development methodology and quality standards
• Context Continuity System: Cross-process context management for consistent AI collaboration
• Quality Gates Framework: 6-stage quality assurance system with automated validation
• AI Governance & Ethics: Comprehensive compliance and ethical AI development guidelines
• Development Workflow: Structured 4-phase development process with built-in quality controls

Development Features

• Automated Context Loading: Python-based bluebook loader with caching system
• Quality Metrics: Comprehensive measurement system covering code, performance, security, and user experience
• Prompt Engineering: Standardized prompt templates and version control system
• Enterprise Standards: 90%+ type coverage, 80%+ test coverage, ≤15 complexity limit
• Performance Targets: API P95 ≤3s, ≥1000 concurrent users, CPU ≤70%

Development Commands

Claude Bluebook System

# Load Claude development context and bluebook
python3 scripts/claude-bluebook-loader.py --auto-load

# Quick start with bluebook
./claude.sh

# Install bluebook globally
./scripts/install-claude-bluebook.sh --user

# Check system status
./status.sh

# Start development services (if available)
./start.sh

Context Management

# Generate Claude context from bluebook
python3 scripts/claude-bluebook-loader.py --auto-load

# Update context cache
python3 scripts/claude-bluebook-loader.py --quiet

# Load specific project context
python3 scripts/claude-bluebook-loader.py --project-root /path/to/project

Bluebook Development

# View bluebook documentation
cat docs/claude-bluebook/README.md

# Access prompt templates
ls prompts/

# View quality standards
cat docs/claude-bluebook/03-quality-system/

# Check development phase
python3 scripts/claude-bluebook-loader.py | grep "Current Development Phase"

Architecture Overview

Claude Bluebook Framework Architecture

• Core Framework: scripts/claude-bluebook-loader.py - Main bluebook management system
• Context Management: .claude-bluebook-cache.json - Persistent context storage
• Documentation System: docs/claude-bluebook/ - Comprehensive development methodology
• Quality Assurance: Built-in 6-stage quality gate system
• Prompt Engineering: Structured template system with version control

Key Framework Components

Claude Bluebook v1.2 System

The framework provides enterprise-grade AI development methodology:

• Context Continuity: Cross-process context management and automatic loading
• Quality Gates: Requirement, Design, Code, Functionality, Performance, Security gates
• Prompt Standards: Structured templates with semantic versioning
• Development Phases: 4-phase structured development process
• Governance Framework: AI ethics and compliance guidelines

Documentation Structure

docs/claude-bluebook/
├── README.md                 # Framework overview and navigation
├── 01-workflow/             # Claude collaboration development workflow
├── 02-prompt-engineering/   # Prompt engineering standards and templates
├── 03-quality-system/       # Quality assurance and measurement system
├── 04-project-structure/    # Project structure and organization standards
├── 05-documentation/        # Documentation-driven development practices
├── 06-ai-integration/       # AI functionality development specifications
├── 07-frontend/            # Frontend development standards
├── 08-backend/             # Backend development standards
├── 09-observability/       # Monitoring and observability systems
├── 10-security/            # Security development standards
├── 11-deployment/          # Deployment and operations specifications
├── 12-team-collaboration/  # Team collaboration and communication
├── 13-maintenance/         # Maintenance and optimization processes
└── governance/             # AI governance and ethical guidelines

Quality Management System

• Code Quality: ≥90% type coverage, ≥80% test coverage, ≤15 complexity limit
• Performance: API P95 ≤3s, ≥1000 concurrent users, CPU ≤70%, memory ≤80%
• Security: <1 vulnerability/KLOC, 100% OWASP compliance, 100% auth coverage
• Documentation: 100% API docs, ≥80% code comments, complete README

Development Workflow

• Phase 1: AI Parsing Engine (Weeks 1-4) - PDF processing and question extraction
• Phase 2: Intelligent Classification (Weeks 5-8) - Vector embedding and knowledge tagging
• Phase 3: AI Tutor System (Weeks 9-12) - Dialog engine and personalized tutoring
• Phase 4: Recommendation System (Weeks 13-16) - User profiling and content recommendation

Project Structure

• Framework Core: scripts/ - Python bluebook loader and management tools
• Documentation: docs/ - Complete development methodology and standards
• Configuration: .claude/ - Claude configuration and system prompts
• Cache System: .claude-bluebook-cache.json - Persistent context and metadata
• Templates: prompts/ - Structured prompt templates organized by category
• Feedback: ai_feedback/ - AI performance analytics and improvement data
• Governance: governance/ - AI ethics and compliance documentation

Development Guidelines

Claude Bluebook Framework Usage

• Context Loading: Always start development sessions with python3 scripts/claude-bluebook-loader.py --auto-load
• Quality Standards: Follow the 6-stage quality gate system for all development work
• Prompt Templates: Use standardized prompt templates from the bluebook system
• Documentation: Maintain documentation-driven development practices
• Version Control: Use semantic versioning for all prompt templates and framework components

Development Workflow Integration

• Phase Management: Track current development phase using the bluebook system
• Quality Gates: Ensure each gate is passed before proceeding to next development stage
• Metrics Tracking: Monitor development metrics through the built-in measurement system
• Context Continuity: Maintain context across Claude sessions using the cache system

Quality Assurance Practices

• Automated Validation: Use built-in quality checks for code, performance, and security
• Metrics Compliance: Ensure all deliverables meet the defined quality standards
• Documentation Standards: Follow the documentation-driven development methodology
• Testing Requirements: Maintain ≥80% test coverage for all code components

AI Collaboration Best Practices

• Clear Requirements: Use structured prompt templates for all development requests
• Context Management: Leverage the context continuity system for consistent collaboration
• Quality Feedback: Provide specific feedback using the bluebook quality framework
• Continuous Improvement: Use the AI feedback system to improve collaboration quality

Performance Considerations

Framework Performance

• Context Loading: Sub-second context generation and caching
• Quality Gate Processing: Automated validation with minimal overhead
• Template Management: Efficient prompt template versioning and retrieval
• Cache System: Persistent context storage for fast session restoration

Development Efficiency Metrics

• Requirement Understanding: ≤30 minutes for comprehensive analysis
• Solution Design: ≤2 hours for complete technical design
• Implementation Speed: ≥50 LOC/hour with quality standards
• Defect Resolution: ≤1 hour for identified issues

Quality Assurance Performance

• Automated Testing: Comprehensive test suite with ≥80% coverage requirement
• Code Review: Automated static analysis with human validation
• Security Scanning: Integrated vulnerability assessment
• Performance Monitoring: Real-time metrics collection and alerting

Deployment

Framework Deployment

• Global Installation: Use ./scripts/install-claude-bluebook.sh --user for system-wide access
• Project Integration: Copy bluebook framework to new projects with context preservation
• Configuration Management: Automated environment detection and setup
• Cache Management: Persistent context storage across sessions and projects

Environment Setup

• Development: Use python3 scripts/claude-bluebook-loader.py --auto-load for optimal setup
• Production: Deploy with ./scripts/install-claude-bluebook.sh for enterprise environments
• CI/CD Integration: Automated context loading and quality validation in pipelines
• Multi-project Support: Context isolation and management for multiple concurrent projects

Key Files to Understand

Framework Core Files

• scripts/claude-bluebook-loader.py - Main bluebook management and context loading system
• .claude-bluebook-cache.json - Persistent context cache and metadata storage
• docs/claude-bluebook/README.md - Complete framework documentation and methodology
• CLAUDE.md - This file - Project-specific guidance and integration instructions

Configuration Files

• .claude/system-prompt.md - Auto-generated Claude context and role definition
• prompts/ - Structured prompt templates organized by development category
• governance/ - AI ethics, compliance, and governance documentation
• ai_feedback/ - Performance analytics and improvement tracking

Documentation Structure

• docs/claude-bluebook/01-workflow/ - Claude collaboration development workflow
• docs/claude-bluebook/02-prompt-engineering/ - Prompt engineering standards and templates
• docs/claude-bluebook/03-quality-system/ - Quality assurance and measurement framework
• docs/claude-bluebook/governance/ - AI governance and ethical development guidelines

Utility Scripts

• start.sh - Development environment initialization (if applicable)
• status.sh - System status and configuration verification
• claude.sh - Quick bluebook loading and context generation

This AI development framework is designed for enterprise-grade software delivery with comprehensive quality assurance, efficient AI collaboration, and seamless integration into existing development workflows.

Claude Bluebook System Details

Framework Version and Copyright

• Current Version: Claude Bluebook v1.2.0
• Release Date: January 18, 2025
• Copyright: © 2025 灵境万象智能科技有限公司 (Lingjing Wanxiang Intelligent Technology Co., Ltd.)
• License: All Rights Reserved

Core Development Principles

1. AI-Driven Development: Claude as primary developer with human guidance and quality control
2. Quality-Built: Every development stage includes explicit quality standards and automated checks
3. Documentation as Contract: All design decisions and interface definitions are documented
4. Iterative Optimization: Continuous improvement of development processes and AI collaboration

Quality Gate System

The framework implements a 6-stage quality gate system:

1. Requirement Understanding Gate

   • ≥95% requirement understanding accuracy
   • ≥90% technical point identification completeness
   • 100% risk assessment completeness

2. Design Solution Gate

   • 100% technical solution feasibility
   • ≥90% architectural design rationality
   • 100% interface design completeness

3. Code Quality Gate

   • 100% code standard compliance
   • 100% static code analysis pass rate
   • ≥80% unit test coverage
   • ≤10 code complexity

4. Functionality Gate

   • 100% functional completeness
   • 100% API interface correctness
   • 100% data consistency
   • ≥95% error handling completeness

5. Performance Gate

   • API P95 response time ≤3s
   • Page load time ≤3s
   • ≥1000 concurrent users
   • ≤80% memory usage

6. Security Gate

   • 100% security vulnerability scan pass rate
   • 100% dependency security check pass rate
   • 100% authentication authorization test pass rate
   • 100% data protection compliance

Development Workflow

The framework follows a structured 4-phase development process:

• Phase 1: AI Parsing Engine (Weeks 1-4)

  • PDF processing pipeline
  • Question extraction and recognition
  • Image and formula processing
  • Performance optimization

• Phase 2: Intelligent Classification (Weeks 5-8)

  • Vector embedding system
  • Knowledge point auto-tagging
  • Similarity matching algorithms
  • Human review interface

• Phase 3: AI Tutor System (Weeks 9-12)

  • Core dialog engine
  • Personalized tutoring logic
  • Multi-modal interaction
  • Progress tracking

• Phase 4: Recommendation System (Weeks 13-16)

  • User profiling system
  • Recommendation algorithms
  • A/B testing framework
  • Performance optimization

Prompt Engineering Standards

The framework includes comprehensive prompt engineering standards:

• Structured Templates: Standardized formats for different types of development tasks
• Version Control: Semantic versioning for prompt templates (v1.0.0, v1.1.0, etc.)
• Quality Metrics: Accuracy, completeness, quality, and efficiency measurements
• Best Practices: Role-setting, context provision, example-driven approaches

Enterprise Integration Features

• Global Installation: System-wide availability with ./scripts/install-claude-bluebook.sh
• Project Templates: Quick setup for new projects with bluebook integration
• CI/CD Integration: Automated context loading and quality validation
• Multi-project Support: Isolated context management for concurrent projects

AI Ethics and Governance

The framework includes comprehensive AI governance:

• Fairness and Non-discrimination: Bias prevention and inclusive design
• Transparency and Explainability: Clear AI decision-making processes
• Privacy Protection: Data protection and user privacy safeguards
• Human Oversight and Control: Human-in-the-loop validation and control mechanisms