Tiling Trees MCP Server

A Model Context Protocol (MCP) server for exploring research ideas using the tiling trees method. This server provides tools to create, organize, and analyze hierarchical research structures through a tile-based approach.

Features

• Hierarchical Research Nodes: Create and organize research ideas as interconnected nodes (tiles)

• Multiple Node Types: Support for questions, hypotheses, observations, methods, results, and insights

• Relationship Mapping: Link nodes with various relationship types (supports, contradicts, extends, etc.)

• Smart Analysis: Identify research gaps, clusters, and patterns in your research tree

• Multiple Export Formats: Export to JSON, Markdown, Mermaid diagrams, and DOT graphs

• Search & Exploration: Query and navigate through your research structure

Installation

Configuration

Add to your MCP settings file (e.g., claude_desktop_config.json):

Available Tools

1. create_research_node

Create a new research node/tile with a concept, question, or idea.

Parameters:

• title (required): Brief title for the research node

• content (required): Detailed content, question, or hypothesis

• parentId (optional): ID of parent node to attach this to

• tags (optional): Array of tags for categorization

• type (optional): Type of node - question, hypothesis, observation, method, result, or insight

Example:

2. split_research_node

Split a complex research node into multiple sub-nodes for better organization.

Parameters:

• nodeId (required): ID of the node to split

• subNodes (required): Array of sub-nodes to create with title, content, and optional type

Example:

3. link_research_nodes

Create relationships between nodes beyond parent-child hierarchy.

Parameters:

• sourceId (required): Source node ID

• targetId (required): Target node ID

• relationshipType (required): Type - supports, contradicts, extends, relates_to, or prerequisite

• notes (optional): Additional notes about the relationship

4. explore_research_path

Explore the research tree from a specific node or view all roots.

Parameters:

• nodeId (optional): Starting node ID (omit to show all root nodes)

• depth (optional): How many levels deep to explore (default: 3)

• includeLinks (optional): Include cross-references (default: true)

5. search_research_tree

Search across all research nodes.

Parameters:

• query (optional): Search text (searches title and content)

• tags (optional): Filter by tags

• type (optional): Filter by node type

6. get_research_insights

Analyze the research tree to identify patterns and opportunities.

Parameters:

• analysisType (required): Type of analysis - gaps, clusters, paths, or summary

• focusArea (optional): Focus on specific tag or node ID

Analysis Types:

• gaps: Find unexpanded nodes, unanswered questions, and hypotheses without methods

• clusters: Identify groups of related nodes by tags and types

• paths: Analyze research paths and find most connected nodes

• summary: Get overall statistics and recent activity

7. update_research_node

Update an existing research node.

Parameters:

• nodeId (required): ID of the node to update

• title (optional): New title

• content (optional): Updated content

• tags (optional): Updated tags

• status (optional): Status - exploring, active, completed, or archived

8. export_research_tree

Export the research tree in various formats.

Parameters:

• format (required): Export format - json, markdown, mermaid, or dot

• nodeId (optional): Export from specific node (default: entire tree)

9. get_research_statistics

Get statistics about the research tree structure.

Usage Examples

Example 1: Building a Research Question Tree

Example 2: Finding Research Gaps

Example 3: Exporting for Visualization

Research Workflow

1. Start with Questions: Create root nodes with research questions

2. Develop Hypotheses: Add potential answers as child nodes

3. Design Methods: Attach methods to test each hypothesis

4. Record Observations: Document what you observe during research

5. Capture Results: Add result nodes with findings

6. Extract Insights: Create insight nodes for key learnings

7. Link Relationships: Connect related concepts across the tree

8. Analyze & Iterate: Use insights tools to find gaps and new directions

Node Types Guide

• question: Research questions to investigate

• hypothesis: Proposed answers or explanations

• observation: Empirical observations or data points

• method: Approaches or procedures to test hypotheses

• result: Outcomes from applying methods

• insight: Key learnings or conclusions

Export Formats

JSON

Complete data structure with all nodes and links.

Markdown

Human-readable hierarchical document with sections for each node.

Mermaid

Flowchart diagram syntax for visualization (use with Mermaid.js).

DOT

GraphViz format for generating publication-quality graphs.

Tips for Effective Research Trees

1. Keep tiles focused: Each node should represent one clear concept

2. Use appropriate types: Distinguish between questions, hypotheses, and results

3. Tag consistently: Use tags to create cross-cutting themes

4. Link generously: Connect related ideas even if not parent-child

5. Regular analysis: Use insights tools to guide your research direction

6. Update status: Mark nodes as exploring, active, completed, or archived

Integration with Tiling Trees Web Interface

This MCP server can work alongside the tiling-trees web interface. Export your research tree and import it into the web interface for visual exploration, or use the MCP server to programmatically build structures that you visualize in the web app.

License

MIT