BuildContext

Explore and traverse concept relationships in your Maenifold knowledge graph through intelligent graph navigation. This tool discovers connections between [[concepts]] by analyzing co-occurrence patterns in your memory files, building contextual understanding through relationship exploration and multi-hop graph traversal.

When to Use This Tool

Key Features

Parameters

ParameterTypeRequiredDescriptionExample
conceptNamestringYesCONCEPT name to build context around (NOT a file!)”Machine Learning”
depthintNoHow many hops in the CONCEPT GRAPH (default: 2)3
maxEntitiesintNoMax entities to return (default: 20)50
includeContentboolNoInclude full content (default: false)true

Usage Examples

Basic Concept Exploration

{
  "conceptName": "Neural Networks"
}

Discovers direct relationships to Neural Networks concept with default 2-hop depth and 20 entity limit.

Focused Direct Relationships

{
  "conceptName": "GraphRAG",
  "depth": 1,
  "maxEntities": 10
}

Explores only immediate (1-hop) relationships to GraphRAG, limiting results to 10 most relevant concepts.

Deep Context Building

{
  "conceptName": "Transformer Architecture", 
  "depth": 3,
  "maxEntities": 50
}

Performs deep 3-hop traversal around Transformer Architecture, returning up to 50 related concepts for comprehensive context.

Research Preparation

{
  "conceptName": "Memory Systems",
  "depth": 2,
  "maxEntities": 30,
  "includeContent": true
}

Builds research context around Memory Systems with 2-hop exploration, returning 30 entities with full content for detailed analysis.

Common Patterns

Research Context Building

Use BuildContext before starting research to understand the conceptual landscape. Discover related topics, competing approaches, and knowledge gaps in your current understanding.

Sequential Thinking Preparation

Build context around key concepts before Sequential Thinking sessions to prime your analysis with related ideas and connection patterns.

Knowledge Gap Discovery

Explore concept neighborhoods to identify missing relationships or underdeveloped areas in your knowledge graph.

Topic Bridge Finding

Use multi-hop traversal to discover conceptual bridges between different domains, revealing unexpected connections in your knowledge.

Workflow Context Gathering

Build comprehensive context before structured workflows like design thinking or problem solving to ensure rich conceptual input.

Concept Clustering Analysis

Analyze how concepts cluster together through co-occurrence patterns, revealing natural knowledge organization in your memory system.

Graph Traversal Patterns

1-Hop Direct Relationships

Perfect for understanding immediate concept connections and finding most relevant related ideas.

{"conceptName": "Deep Learning", "depth": 1}

2-Hop Extended Context

Ideal balance of breadth and focus, discovering concept neighborhoods while maintaining relevance.

{"conceptName": "Attention Mechanism", "depth": 2}

3+ Hop Broad Exploration

Use for discovering distant relationships and finding unexpected conceptual bridges across domains.

{"conceptName": "Optimization", "depth": 3, "maxEntities": 40}

Focused vs. Broad Entity Limits

Troubleshooting

Error: “CONCEPT ‘X’ not found in graph”

Cause: The concept doesn’t exist in your knowledge graph or needs different spelling
Solution: Run Sync to update graph from recent memory files, or check concept spelling and capitalization

Error: “Run sync first”

Cause: Knowledge graph is empty or outdated compared to memory files
Solution: Execute Sync tool to extract concepts from memory files and build/update the graph database

No relationships found for existing concept

Cause: Concept exists but appears in isolation without co-occurring with other [[concepts]]
Solution: Add the concept to existing memory files alongside related concepts to build relationships

Traversal depth returns limited results

Cause: Your knowledge graph has sparse connections or concept clusters are isolated
Solution: Increase maxEntities parameter or create memory files that bridge concept areas with shared [[concepts]]

Performance issues with high depth/entity limits

Cause: Very dense concept graphs with high connectivity can create large result sets
Solution: Reduce depth to 1-2 or lower maxEntities to 20-30 for faster traversal

Missing expected concept relationships

Cause: Related concepts may exist in different memory files without co-occurrence
Solution: Create bridging memory files that mention related concepts together to establish graph connections

Graph Theory Foundation

BuildContext implements LazyGraphRAG principles through:

Co-Occurrence Relationship Model

Concepts that appear together in the same memory file create weighted edges based on frequency of co-occurrence across multiple files.

Breadth-First Traversal

Explores concept relationships level by level, prioritizing closer relationships while extending to discover broader context.

Weighted Edge Ranking

Relationships are ranked by co-occurrence count, ensuring most frequent concept pairings appear first in results.

Source Provenance

Every relationship tracks which memory files contain the co-occurring concepts, enabling verification and deeper investigation.

Integration with Maenifold Architecture

Sequential Thinking Integration

Use BuildContext to gather conceptual context before thinking sessions:

BuildContext → SequentialThinking (with rich concept context)

Workflow Preparation

Build comprehensive context before structured methodologies:

BuildContext → Design Thinking Workflow (with related concepts)

Knowledge Graph Evolution

As you create more memory files with [[concepts]], BuildContext discoveries become richer and more connected.

Memory System Synergy

BuildContext reveals the structure of your knowledge while SearchMemories finds the content - complementary tools for knowledge exploration.

Ma Protocol Compliance

BuildContext follows Maenifold’s Ma Protocol principles:

This tool transforms your memory files into an explorable knowledge graph, revealing the hidden connections and patterns in your accumulated knowledge through intelligent relationship traversal.