The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
MCP N8n Builder
MCP server for programmatic creation and management of n8n workflows. Enables AI assistants to build, modify, and manage workflows without direct user intervention through a…
Install
git clone https://github.com/spences10/mcp-n8n-builderInstall command taken from the project's README or registry entry. Always review code before running it.
README excerpt
Read full README on GitHub →mcp-n8n-builder
A Model Context Protocol (MCP) server for programmatic creation and management of n8n workflows. This server provides tools and resources for interacting with n8n's REST API, allowing AI assistants to build, modify, and manage workflows without direct user intervention.
Features
- 🔄 Workflow Management: Create, read, update, delete, activate, and deactivate workflows
- 📊 Execution Management: List and retrieve workflow execution details
- ✅ Schema Validation: Comprehensive validation with Zod for both input and output data
- 🔍 Node Validation: Validates node types against n8n's available nodes before workflow creation to prevent errors
- 🤔 Smart Suggestions: Provides suggestions for similar node types when invalid nodes are detected
- 🛠️ Error Handling: Detailed error messages for troubleshooting
- 📝 Documentation: Extensive documentation for all tools and resources
- 🔍 Search & Filter: Filter workflows by name, tags, or active status
⚠️ Token Consumption Warning
This tool is a "token monster" due to the nature of n8n workflows:
- N8N workflows are complex JSON structures with nested nodes, connections, parameters, and settings
- A single workflow can easily consume thousands of tokens when viewed or edited
- The JSON structure cannot be meaningfully reduced without losing essential workflow information
- LLM context windows are finite - complex workflows may exceed available context
Mitigation Strategies
The tool implements several optimizations to reduce token usage:
- Verbosity Control: Use
OUTPUT_VERBOSITY=concise(default) for summaries instead of full JSON - Object Simplification: Workflow listings show only essential fields (id, name, status, node count)
- Selective Guide Inclusion: Error messages include only relevant sections of the composition guide
- Per-tool Verbosity: Override verbosity on individual tool calls with the
verbosityparameter
Recommended Usage
- List workflows first to identify what you need (low token cost)
- Work with simple workflows when possible
- Use
conciseverbosity unless you specifically need full JSON - Break complex workflows into smaller pieces for easier management
- Consider the token cost before retrieving or editing large workflows
Configuration
This server requires configuration through your MCP client. Here are examples for different environments:
Cline Configuration
Add this to your Cline MCP settings:
{
"mcpServers": {
"n8n-workflow-builder": {
"command": "npx",
"args": ["-y", "mcp-n8n-builder"],
"env": {
"N8N_HOST": "http://localhost:5678/api/v1",
"N8N_API_KEY": "your-n8n-api-key",
"OUTPUT_VERBOSITY": "concise" // Options: 'concise' or 'full'
}
}
}
}
Claude Desktop with WSL Configuration
For WSL environments, add this to your Claude Desktop configuration:
{
"mcpServers": {
"n8n-workflow-builder": {
"command": "wsl.exe",
"args": [
"bash",
"-c",
"N8N_HOST=http://localhost:5678/api/v1 N8N_API_KEY=your-n8n-api-key OUTPUT_VERBOSITY=concise npx -y mcp-n8n-builder"
]
}
}
}
Environment Variables
The server can be configured using environment variables:
Similar plugs
Reverse engineer anything with agents, from app behavior down to native binaries.
Build Agentic workflows, RAG pipelines, with rich AI model and tool support on one collaborative workspace. Deploy on cloud, VPC, or self-hosted, so teams move from prototype…
A cross-platform desktop All-in-One assistant for Claude Code, Codex, OpenCode, OpenClaw, Grok Build & Hermes Agent. Only official website: ccswitch.io