Branch Thinking

🚦 Branch-Thinking MCP Tool

What’s New (2025-04):

  • Advanced visualization: clustering (k-means/degree), centrality overlays, edge bundling, and agentic overlays for tasks and priorities
  • Agentic cache & prefetch: LRU+TTL caches for embeddings, summaries, analytics, and proactive agent cache warming
  • Enhanced analytics: real-time, multi-branch, and focusNode support; agent-optimized metadata
  • Upgraded documentation and onboarding for agents and users

Features

  • 🌳 Branch Management: Create, focus, and navigate multiple lines of thought
  • 🔗 Cross-References: Link related thoughts across branches (typed, scored)
  • 💡 AI Insights: Automatic insight and summary generation
  • 🧠 Semantic Search: Find related thoughts using embeddings
  • 📊 Advanced Visualization:
    • Node clustering (k-means/degree)
    • Centrality overlays (closeness, betweenness)
    • Edge bundling
    • Task overlays (status, priority, next-action)
    • Agentic overlays and metadata for all nodes/edges
    • FocusNode and multi-branch visualization
  • Agentic Cache & Prefetch:
    • LRU+TTL caches for embeddings, summaries, analytics
    • Proactive cache warming for agent workflows
  • 🗂️ Persistent Storage: Queryable, extensible, and never lose a thought
  • 🔄 Real-Time & Multi-Branch: Visualize and analyze multiple branches and nodes in real-time
  • 🛠️ Production-Grade: Robust error handling, performance optimizations, and agent/human-friendly APIs

🛠️ Technology Stack

  • Node.js (18+)
  • TypeScript (4.x)
  • @dagrejs/graphlib: Graph structure, algorithms, and analytics
  • ml-kmeans: Clustering for visualization
  • lru-cache: LRU+TTL caching for embeddings, summaries, analytics
  • @xenova/transformers: Embedding and summarization pipelines
  • @modelcontextprotocol/sdk: MCP protocol and agent integration
  • chalk: CLI output styling
  • Mermaid: Gantt/roadmap visualization
  • pnpm: Fast dependency management

Roadmap (Gantt)


Table of Contents


Why Branch-Thinking?

  • Agentic by Design: Built for both human and AI workflows—every command is agent-friendly.
  • True Branching: Organize, cross-link, and reason over ideas, code, and tasks in parallel.
  • AI-Native: Semantic search, auto-summarization, and insight generation out-of-the-box.
  • Persistent & Visual: Never lose a thought—everything is queryable, visualizable, and extensible.

Summary

Branch-Thinking MCP Tool is an advanced agentic platform for managing, visualizing, and reasoning over branching thoughts, tasks, code, and knowledge. It empowers both AI agents and humans to organize complex projects, cross-link ideas, and automate insight generation using a powerful branch-based paradigm. With semantic search, visualization, and persistent task/code management, it is designed for next-generation collaborative and autonomous workflows.

Branch-Thinking MCP Tool is an advanced agentic platform for managing, visualizing, and reasoning over branching thoughts, tasks, code, and knowledge. It empowers both AI agents and humans to organize complex projects, cross-link ideas, and automate insight generation using a powerful branch-based paradigm. With semantic search, visualization, and persistent task/code management, it is designed for next-generation collaborative and autonomous workflows.


Architecture & Flow


Quick Start

Get up and running in seconds:

pnpm install # Recommended for speed (or npm install) pnpm build node dist/index.js --help # See available commands

Getting Started

1. Clone & Install

git clone https://github.com/your-org/branch-thinking-mcp.git cd branch-thinking-mcp pnpm install # Or npm install pnpm build # Or npm run build

2. Configure (Optional)

For Claude Desktop integration, add to your claude_desktop_config.json:

"branch-thinking": { "command": "node", "args": [ "/your-custom-mcp-dir-here/branch-thinking/dist/index.js" ] }

3. Run

node dist/index.js

Real-World Usage Recipes

1. Knowledge Capture & Linking

# Batch capture meeting notes add-thought dev "Discussed semantic search improvements" note add-thought dev "Agreed to refactor API" decision # Link related thoughts link-thoughts t1 t2 supports "API refactor supports search improvements"

2. Agentic Task Extraction

# Extract and manage tasks from a research branch extract-tasks research list-tasks research open update-task-status task-1 in_progress

3. Visualization for Insight

# Generate and interpret a knowledge graph visualize dev # Review AI-generated summary summarize-branch dev

🧑‍💻 Live Example: Agentic Workflow

# 1. Create a new branch for your project or idea create-branch "AI Research" # 2. Add thoughts and observations add-thought [branchId] "Explore semantic search for agent workflows" analysis add-thought [branchId] "Test cross-linking and summarization" observation # 3. Link related thoughts link-thoughts [thoughtId1] [thoughtId2] supports "Thought 2 validates Thought 1" # 4. See your knowledge graph visualize [branchId] # 5. Extract tasks and get AI review extract-tasks [branchId] review-branch [branchId]

Replace [branchId] and [thoughtIdX] with actual IDs from list and history.


Command Reference

Branch Management

CommandDescription
listShow all branches with status
focus [branchId]Switch focus to a branch
history [branchId?]Show thought history
summarize-branch [branchId?]AI summary of branch
review-branch [branchId?]AI review of branch
visualize [branchId?]Visual graph of connections

Thought & Insight Management

CommandDescription
insights [branchId?]Get AI-generated insights
crossrefs [branchId?]Show cross-references
hub-thoughts [branchId?]List hub thoughts
semantic-search [query]Find similar thoughts
link-thoughts [from] [to] [type] [reason?]Link two thoughts
add-snippet [content] [tags]Save a code snippet
snippet-search [query]Search code snippets
doc-thought [thoughtId]Document a thought

Task Management

CommandDescription
extract-tasks [branchId?]Extract actionable items
list-tasks [branchId] [status] [assignee] [due]List/filter tasks
update-task-status [taskId] [status]Update a task’s status
summarize-tasks [branchId]Summarize tasks

AI & Knowledge

CommandDescription
ask [question]AI answer from knowledge base

Best Practices

  • Always start with create-branch to ensure clean context.
  • Use list and focus to navigate between projects or lines of thought.
  • Leverage summarize-branch and insights after adding several thoughts to get AI-generated context.
  • Use link-thoughts to explicitly connect ideas, tasks, or code for richer semantic graphs.
  • After code changes, always run pnpm lint and pnpm build to catch errors early.
  • Decompose complex goals into sequences of thought/task/insight commands.
  • Iterate and adapt: Use feedback from summaries, reviews, and visualizations to refine next actions.
  • Explicitly specify parameters (branchId, status, assignee, etc.) for precise results.
  • Use cross-references and multi-hop links to foster creativity and bridge ideas.
  • Prompt agents (Claude, GPT-4, etc.) to "think step by step" or "use chain of thought" for best results.

Security

  • All persistent data is stored locally (default: project directory or MCP_STORAGE_PATH)
  • No external API calls unless configured
  • Agents/users are responsible for privacy of stored thoughts and tasks
  • To report security issues, please open an issue or email the maintainer.

Troubleshooting and FAQ

Q: The tool isn't responding! A: Check the MCP server logs and ensure configuration is correct.

Q: How do I reset storage? A: Delete or move the persistent storage directory (see config).

Q: How do I add a new command? A: Extend handleCommand in src/index.ts and document it in the README.

Accessibility and Internationalization

  • All badges/images have descriptive alt text.
  • English is the default language; contributions for translations are welcome.
  • Please open a PR or issue if you want to help localize this tool.

Contributing

Contributions, issues, and feature requests are welcome! Please open a PR or issue on GitHub.

  1. Fork this repo
  2. Create a new branch (git checkout -b feature/your-feature)
  3. Commit your changes
  4. Push to the branch
  5. Open a Pull Request

Credits

  • Concept & Testing: @ssdeanx
  • Core Code Generation: Claude, GPT-4, and Cascade
  • Implementation, Fixes, and Documentation: @ssdeanx

License

MIT

-
security - not tested
A
license - permissive license
-
quality - not tested

local-only server

The server can only run on the client's local machine because it depends on local resources.

An MCP server that enables managing multiple lines of thought with features like branch navigation, cross-references between related thoughts, and insight generation from key points.

  1. Features
    1. 🛠️ Technology Stack
      1. Roadmap (Gantt)
        1. Table of Contents
          1. Why Branch-Thinking?
            1. Summary
              1. Architecture & Flow
                1. Quick Start
                  1. Getting Started
                    1. 1. Clone & Install
                    2. 2. Configure (Optional)
                    3. 3. Run
                  2. Real-World Usage Recipes
                    1. 1. Knowledge Capture & Linking
                    2. 2. Agentic Task Extraction
                    3. 3. Visualization for Insight
                  3. 🧑‍💻 Live Example: Agentic Workflow
                    1. Command Reference
                      1. Branch Management
                      2. Thought & Insight Management
                      3. Task Management
                      4. AI & Knowledge
                    2. Best Practices
                      1. Security
                        1. Troubleshooting and FAQ
                          1. Accessibility and Internationalization
                            1. Contributing
                              1. References and Related Projects
                                1. Credits
                                  1. License

                                    Related MCP Servers

                                    • A
                                      security
                                      A
                                      license
                                      A
                                      quality
                                      An MCP server for navigating thought processes using branches, supporting thought cross-references and priority tracking to enhance insight generation and structured idea exploration.
                                      Last updated -
                                      1
                                      20
                                      TypeScript
                                      MIT License
                                    • A
                                      security
                                      A
                                      license
                                      A
                                      quality
                                      An adaptation of the MCP Sequential Thinking Server designed to guide tool usage in problem-solving. This server helps break down complex problems into manageable steps and provides recommendations for which MCP tools would be most effective at each stage.
                                      Last updated -
                                      1
                                      361
                                      121
                                      TypeScript
                                      MIT License
                                    • -
                                      security
                                      A
                                      license
                                      -
                                      quality
                                      An MCP server that implements the 'think' tool, providing Claude with a dedicated space for structured thinking during complex problem-solving tasks to improve reasoning capabilities.
                                      Last updated -
                                      48
                                      Python
                                      MIT License
                                      • Linux
                                      • Apple

                                    View all related MCP servers

                                    ID: 6dfti7wyh5