diff --git a/project_context.md b/project_context.md
new file mode 100644
index 0000000..7af71e6
--- /dev/null
+++ b/project_context.md
@@ -0,0 +1,163 @@
+# KnowledgeBox Project Context
+
+## Project Vision
+
+KnowledgeBox is a collaborative knowledge-gathering platform that combines the familiar interface of document editors (like Google Docs) with AI-powered learning assistance. Users can create topic-specific "KnowledgeBoxes" where they work alongside an AI agent to systematically explore, organize, and understand complex subjects.
+
+## Core Concept
+
+A KnowledgeBox is a living document that evolves as users and AI agents collaboratively:
+- Research and gather information about a specific topic
+- Structure knowledge in a logical, learnable format
+- Identify and fill knowledge gaps
+- Create comprehensive learning resources
+
+## Key Features
+
+### 1. Collaborative Knowledge Building
+- **User-AI Partnership**: Users and AI agents work together in real-time
+- **Topic-Focused**: Each KnowledgeBox centers on a specific subject area
+- **Iterative Learning**: Content evolves through continuous collaboration
+
+### 2. Intelligent Content Organization
+- **Automatic Structuring**: AI helps organize information into logical sections
+- **Knowledge Mapping**: Visual representation of topic relationships
+- **Gap Identification**: AI identifies missing or underdeveloped areas
+
+### 3. Interactive Learning Experience
+- **Guided Exploration**: AI suggests learning paths and next steps
+- **Question Generation**: AI poses relevant questions to deepen understanding
+- **Resource Recommendations**: AI suggests additional learning materials
+
+## User Experience Flow
+
+### 1. KnowledgeBox Creation
+```
+User → Creates new KnowledgeBox → Defines topic/learning goal → AI initializes structure
+```
+
+### 2. Collaborative Content Development
+```
+User adds initial knowledge → AI suggests improvements/expansions → User refines → AI identifies gaps → Cycle continues
+```
+
+### 3. Knowledge Refinement
+```
+AI organizes content → Suggests connections → Identifies missing areas → User fills gaps → AI validates and enhances
+```
+
+## AI Agent Responsibilities
+
+### Content Organization
+- Structure information into logical hierarchies
+- Create topic outlines and learning paths
+- Suggest content reorganization for better flow
+
+### Knowledge Gap Analysis
+- Identify missing fundamental concepts
+- Suggest areas for deeper exploration
+- Recommend prerequisite knowledge
+
+### Learning Enhancement
+- Generate relevant questions and exercises
+- Provide examples and analogies
+- Suggest practical applications
+
+### Research Assistance
+- Find relevant resources and references
+- Summarize complex information
+- Validate information accuracy
+
+## Example Use Cases
+
+### 1. PLC Programming Learning
+**Initial State**: User creates "PLC Programming Fundamentals" KnowledgeBox
+**AI Response**: 
+- Creates basic outline (hardware, software, programming languages, applications)
+- Suggests starting with basic concepts
+- Identifies need for hands-on examples
+
+**Collaboration Flow**:
+- User adds notes about ladder logic
+- AI suggests expanding with timing diagrams
+- AI identifies missing topic: I/O configuration
+- User researches I/O, AI helps organize findings
+- AI suggests practical exercises
+
+### 2. Data Science Concepts
+**Initial State**: User creates "Machine Learning for Beginners" KnowledgeBox
+**AI Response**:
+- Structures learning path from statistics to advanced algorithms
+- Identifies mathematical prerequisites
+- Suggests practical project ideas
+
+### 3. Historical Research
+**Initial State**: User creates "World War II Pacific Theater" KnowledgeBox
+**AI Response**:
+- Creates chronological and thematic organization
+- Identifies key figures, battles, and decisions
+- Suggests primary and secondary sources
+
+## Technical Requirements
+
+### Core Functionality
+- **Real-time Collaboration**: Seamless user-AI interaction
+- **Document Editing**: Rich text editing with collaborative features
+- **Knowledge Structuring**: Hierarchical organization tools
+- **Search and Navigation**: Easy content discovery within KnowledgeBoxes
+
+### AI Integration
+- **Natural Language Processing**: Understanding user intent and content
+- **Knowledge Reasoning**: Identifying logical connections and gaps
+- **Content Generation**: Creating structured outlines and suggestions
+- **Learning Analytics**: Tracking progress and identifying areas for improvement
+
+### User Interface
+- **Intuitive Design**: Familiar document editing experience
+- **AI Interaction**: Clear indicators of AI suggestions and contributions
+- **Visual Organization**: Mind maps, outlines, and knowledge graphs
+- **Progress Tracking**: Visual indicators of learning progress
+
+## Success Metrics
+
+### User Engagement
+- Time spent in KnowledgeBoxes
+- Number of AI-user interactions
+- Content creation and refinement cycles
+
+### Learning Effectiveness
+- Knowledge retention (through built-in assessments)
+- Completion of learning objectives
+- User satisfaction with knowledge organization
+
+### AI Performance
+- Accuracy of gap identification
+- Relevance of suggestions
+- Quality of content organization
+
+## Development Principles
+
+### 1. User-Centric Design
+- The user maintains control over their learning journey
+- AI enhances rather than replaces human curiosity and creativity
+- Clear distinction between user and AI contributions
+
+### 2. Adaptive Learning
+- AI adapts to user's learning style and pace
+- Personalized suggestions based on user background
+- Flexible content organization options
+
+### 3. Knowledge Quality
+- Emphasis on accurate, well-sourced information
+- Validation of AI suggestions and content
+- Encouragement of critical thinking and verification
+
+## Future Vision
+
+KnowledgeBox aims to become the go-to platform for:
+- Students researching complex topics
+- Professionals learning new skills
+- Researchers organizing domain knowledge
+- Anyone seeking to systematically understand a subject
+
+The platform should feel like having a knowledgeable research partner who helps you think through problems, organize information, and discover new connections in your learning journey.
\ No newline at end of file