DecisionNode MCP

v1.0.0Knowledge & Memorystable

CLI + Local MCP - A shared structured memory store across Claude Code, Cursor, Windsurf, Antigravity, and every MCP client. Semantically queryable.

agentic-workflowaianti-gravityarchitectural-decisionsclaude
Share:
38
Stars
0
Downloads
0
Weekly
0/5

What is DecisionNode MCP?

DecisionNode MCP is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to cli + local mcp - a shared structured memory store across claude code, cursor, windsurf, antigravity, and every mcp client. semantically queryable.

CLI + Local MCP - A shared structured memory store across Claude Code, Cursor, Windsurf, Antigravity, and every MCP client. Semantically queryable.

This server falls under the Knowledge & Memory category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • CLI + Local MCP - A shared structured memory store across Cl

Use Cases

Development decision storage
Semantic query of decisions
Cross-tool memory
decisionnode

Maintainer

LicenseMIT
Languagetypescript
Versionv1.0.0
UpdatedMay 17, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx decisionnode

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use DecisionNode MCP

DecisionNode is a local CLI and MCP server that provides a shared, semantically queryable memory store for architectural and development decisions across all your AI coding tools — including Claude Code, Cursor, Windsurf, and any other MCP client. Decisions are stored locally, embedded with Google Gemini's free-tier embedding API for semantic search, and queryable by natural language so an AI assistant can retrieve the rationale behind past choices before making new ones. It prevents AI agents from contradicting or undoing earlier architectural decisions by giving them persistent, searchable context.

Prerequisites

  • Node.js 18 or higher installed
  • npm package manager
  • A free Google Gemini API key (used for semantic embeddings — obtain at aistudio.google.com)
  • An MCP-compatible client such as Claude Desktop, Claude Code, or Cursor
1

Install DecisionNode globally

Install the decisionnode package globally from npm. This makes both the decide CLI and the decide-mcp server available system-wide.

npm install -g decisionnode
2

Initialize a project decision store

Navigate to your project directory and initialize a local decision store. This creates the .decisionnode/ directory that holds your decisions and embeddings.

cd your-project
decide init
3

Configure the Gemini API key

Run the setup wizard to enter your free Google Gemini API key. This key is used to generate semantic embeddings for your stored decisions, enabling similarity search.

decide setup
4

Add the MCP server to your AI client

Register the DecisionNode MCP server with Claude Code using the mcp add command. This makes the nine MCP tools available to Claude in any project session.

claude mcp add decisionnode -s user decide-mcp
5

Store your first decision

Add an architectural decision using the CLI. Include a subject (category) and description (rationale). Decisions are immediately embedded and searchable.

decide add -s Backend -d "Use PostgreSQL over MongoDB — data is relational and we need ACID guarantees for billing"
6

Search decisions semantically

Query stored decisions by meaning rather than exact keywords. DecisionNode returns the most relevant decisions based on embedding similarity.

decide search "database choice for billing"

DecisionNode MCP Examples

Client configuration

claude_desktop_config.json entry for the DecisionNode MCP server using the globally installed decide-mcp command.

{
  "mcpServers": {
    "decisionnode": {
      "command": "decide-mcp",
      "args": []
    }
  }
}

Prompts to try

Example prompts to leverage DecisionNode's decision memory from within Claude or another AI coding assistant.

- "Before changing the auth system, search DecisionNode for any past decisions about authentication"
- "Add a decision: we chose Tailwind over CSS Modules because team velocity was higher"
- "List all architectural decisions tagged under the Frontend subject"
- "Did we document why we avoided Redux? Search decisions for state management rationale"
- "Show the full history of the database decision including any updates"

Troubleshooting DecisionNode MCP

decide-mcp command not found after installation

Ensure your npm global bin directory is on your PATH. Run `npm bin -g` to find the directory and add it to your shell profile (e.g., export PATH="$(npm bin -g):$PATH" in ~/.zshrc).

Semantic search returns irrelevant results

Run `decide check` to verify that all decisions have valid embeddings. If embeddings are missing or corrupted, run `decide embed` to regenerate them using your Gemini API key.

Gemini API key rejected during setup

Ensure you are using a key from Google AI Studio (aistudio.google.com), not a Google Cloud Vertex AI key. The free-tier AI Studio key is what DecisionNode expects.

Frequently Asked Questions about DecisionNode MCP

What is DecisionNode MCP?

DecisionNode MCP is a Model Context Protocol (MCP) server that cli + local mcp - a shared structured memory store across claude code, cursor, windsurf, antigravity, and every mcp client. semantically queryable. It connects AI assistants to external tools and data sources through a standardized interface.

How do I install DecisionNode MCP?

Follow the installation instructions on the DecisionNode MCP GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with DecisionNode MCP?

DecisionNode MCP works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is DecisionNode MCP free to use?

Yes, DecisionNode MCP is open source and available under the MIT license. You can use it freely in both personal and commercial projects.

Browse More Knowledge & Memory MCP Servers

Explore all knowledge & memory servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "decisionnode": { "command": "npx", "args": ["-y", "decisionnode"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use DecisionNode MCP?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides