Obsidian Elite RAG MCP Server

# Elite Knowledge Architecture ## Core Design Principles ### 1. Hierarchical Organization (FPEF-Verified) ``` 00-Core/ # Foundational, timeless knowledge 01-Projects/ # Active work with clear outcomes 02-Research/ # Learning and exploration areas 03-Workflows/ # Reusable processes and patterns 04-AI-Paired/ # Claude Code interactions and outputs 05-Resources/ # External references and materials 06-Meta/ # Knowledge about the system itself 07-Archive/ # Historical knowledge (timestamped) 08-Templates/ # Reusable note structures 09-Links/ # External connections and references ``` ### 2. Linking Strategy **Semantic Linking Patterns:** - `[[Concept]]` - Primary relationship - `[[Concept#Subtopic]]` - Granular targeting - `[[Concept|Alias]]` - Alternative naming - `[[Concept]]` - Bidirectional linking - `[[]]` - Placeholder for future connections **Contextual Link Types:** - `→` (Flow): Sequential progression - `⇢` (Suggests): Potential connection - `⇒` (Implies): Logical implication - `⊕` (Combines): Integration point - `⊖` (Contradicts): Tension point ### 3. Tagging System **Hierarchical Tags:** ``` #domain/tech/ai/ml #domain/workflow/automation #domain/research/psychology #status/active/learning #status/archived/complete #priority/critical/urgent #context/work/project-x #method/rag/semantic-search #method/workflow/kanban ``` **Metadata Tags:** ``` #type/concept #type/process #type/resource #type/template #type/question #type/insight ``` ### 4. Note Naming Convention **Format:** `YYYY-MM-DD - Descriptive Title [Context]` **Examples:** - `2024-10-26 - Advanced RAG Techniques [AI/ML]` - `2024-10-26 - Project Phoenix Architecture [Work/Active]` - `2024-10-26 - Learning: Graph Databases [Research]` **Special Prefixes:** - `!` - Action items: `!2024-10-26 - Review Documentation` - `?` - Questions: `?2024-10-26 - How to Optimize Vector Search` - `*` - Important: `*2024-10-26 - Core Architecture Decision` - `+` - Insights: `+2024-10-26 - Connection Between X and Y` ### 5. Content Structure Templates **Standard Note Structure:** ```markdown # Title **Context:** Brief description of this note's purpose and relationships **Last Updated:** 2024-10-26 **Status:** #status/active/learning ## Summary One-sentence summary and key takeaway ## Core Content Main body with structured sections ## Connections - Related: [[Concept A]], [[Concept B]] - Contradicts: [[Concept C]] - Implies: [[Concept D]] ## Actions - [ ] Follow-up on X - [ ] Review Y ## Questions - How does this relate to Z? ## Metadata - **Tags:** #domain/tech #method/rag - **Created:** 2024-10-26 - **Modified:** 2024-10-26 ``` ### 6. Knowledge Graph Optimization **Link Density Principles:** - Each concept should have 3-7 primary connections - Avoid "orphan" notes (no incoming links) - Create "hub" notes for central concepts - Use MOCs (Maps of Content) for topic organization **MOC Structure:** ```markdown # MOC: Artificial Intelligence ## Core Concepts - [[Machine Learning]] - [[Neural Networks]] - [[Natural Language Processing]] ## Applications - [[Computer Vision]] - [[Recommendation Systems]] - [[RAG Systems]] ## Research Areas - [[Transformer Architecture]] - [[Vector Databases]] - [[Knowledge Graphs]] ## Related MOCs - [[MOC: Data Science]] - [[MOC: Software Engineering]] ``` ### 7. Contextual Layering **Layer 1: Core Knowledge** - Fundamental concepts and definitions - Timeless principles and theories - Domain foundations **Layer 2: Applied Knowledge** - Implementation details and examples - Practical applications and case studies - Process documentation **Layer 3: Meta-Knowledge** - Learning journeys and insights - Questions and exploration areas - Personal reflections and synthesis **Layer 4: Connections** - Cross-domain relationships - Pattern recognition - Higher-level abstractions ### 8. Retrieval Optimization **Search-Friendly Content:** - Clear, descriptive titles - Consistent terminology - Keyword optimization - Context-rich summaries **Indexing Strategy:** - Use tags for broad categorization - Use links for specific relationships - Use aliases for terminology variations - Use properties for structured metadata This architecture ensures optimal RAG performance by creating a richly interconnected, hierarchically organized knowledge base that supports multiple retrieval strategies and context synthesis.