atlas-mcp-server
by cyanheads·★ 475·Score 48
A Neo4j-powered MCP server implementing three-tier project/task/knowledge management for LLM agents with deep research capabilities.
Overview
ATLAS is a comprehensive MCP server built on a three-node architecture (Projects, Tasks, Knowledge) that provides structured task management for LLM agents. It leverages Neo4j for graph database capabilities, enabling sophisticated relationship tracking between entities, dependency management, and unified search functionality. The server offers both standard I/O and HTTP transport modes, with experimental web UI for visual management. With 475 stars and recent activity, it appears to be a well-maintained, production-ready solution for complex workflow management.
Try asking AI
After installing, here are 6 things you can ask your AI assistant:
When to choose this
Choose ATLAS when you need a structured knowledge graph for complex project management with dependency tracking for LLM agents.
When NOT to choose this
Avoid ATLAS if you need a lightweight solution without external database dependencies or if you prefer open-source relational databases over Neo4j.
Tools this server exposes
12 tools extracted from the READMEcreate_projectCreate a new project with details like name, description, and status
update_projectUpdate an existing project's properties
create_taskCreate a new task under a specific project
update_taskUpdate an existing task's properties
create_knowledgeAdd a new knowledge item to a project
searchSearch across projects, tasks, and knowledge items
get_projectRetrieve details of a specific project
get_taskRetrieve details of a specific task
get_knowledgeRetrieve details of a specific knowledge item
list_projectsList all projects with optional filtering
list_tasksList tasks with optional filtering by project or status
list_knowledgeList knowledge items with optional filtering
Comparable tools
Installation
Installation
- Clone the repository:
git clone https://github.com/cyanheads/atlas-mcp-server.git
cd atlas-mcp-server- Install dependencies:
npm install- Configure Neo4j (start with Docker):
docker-compose up -dUpdate .env with Neo4j connection details.
- Build the project:
npm run buildClaude Desktop Configuration
Add to claude_desktop_config.json:
{
"mcpServers": {
"atlas": {
"command": "node",
"args": ["dist/index.js"],
"env": {
"NEO4J_URI": "bolt://localhost:7687",
"NEO4J_USER": "neo4j",
"NEO4J_PASSWORD": "yourpassword"
}
}
}
}FAQ
- What database does ATLAS use?
- ATLAS uses Neo4j as its graph database since version 2.0. Previous versions used SQLite.
- Can I use ATLAS without Neo4j?
- No, Neo4j is required for version 2.0 and later. You can self-host with Docker or use Neo4j AuraDB cloud service.
- What transport methods are supported?
- ATLAS supports both Standard I/O (stdio) for local clients and HTTP transport for remote access or web-based integrations.
Compare atlas-mcp-server with
Last updated · Auto-generated from public README + GitHub signals.