Intelligent MCP Orchestrator

A Cognitive Dispatcher that routes tasks between internal MCP tools and external AI providers, with Claude-first bias and self-learning capabilities.

960+ Tests Passing 16 Modules Complete Real API Integration 4 Grafana Dashboards
View Demos AI Assistants MCP Router GitHub

Key Features

Capability-Based Routing

Match tasks to tools via capability tags, not exact names. Decouples task from implementation.

Claude-First Bias

Prefer Anthropic Claude for external calls with 1.5x weight. Fallback to other providers when needed.

Internal Tools First

Try local MCP tools before external providers. Lower latency, no costs, more control.

Retry & Fallback

Exponential backoff with automatic fallback chains. Resilient to transient failures.

Learning Layer

LanceDB for semantic similarity, Neo4j for relationships. Improves tool selection over time.

Full Observability

Prometheus metrics, Grafana dashboards, JSON structured logs. Complete visibility.

6-Module Architecture

M1
Core Orchestrator
89 tests
M2
Config Engine
106 tests
M3
Observability
110 tests
M4
Dashboard
136 tests
M5
Learning
107 tests
M6
Infrastructure
Docker

Extended Modules

projects02: AI Assistant Integration

Multi-provider routing for OpenAI, Gemini, and Claude with health monitoring and feedback collection. 189 tests.

projects03: MCP Meta-Router

Route tasks to external MCP servers (browser-mcp, paaf, research-agent) with unified tool discovery. 226 tests.

Dashboard & Observability

Full visibility into system behavior with real-time monitoring and historical analysis.

4 Grafana Dashboards

MCP Overview, M5 Learning Layer, API Endpoints, and Logs Explorer. Real-time metrics and visualizations.

Learning Layer Visibility

Monitor LanceDB embeddings and Neo4j graph nodes in real-time. See tool success rates and relationships grow.

Prometheus Metrics

15+ custom metrics for learning layer health, vector store stats, and graph store relationships.

REST API Stats

/learning/stats and /learning/health endpoints for programmatic access to learning layer statistics.

Integration Ready

Connect your systems to the orchestrator via REST API or integrate with existing frameworks.

SPINE Integration

Works with SPINE multi-agent orchestration framework for Plan→Execute→Verify→Commit workflows.

MCP Server Compatible

Routes to any MCP server: browser-mcp, paaf, research-agent, filesystem, and custom servers.

REST API Gateway

FastAPI-based dashboard with OpenAPI docs. Call /execute, /tools, /config, and /learning endpoints.

Quick Example

from mcp_core.orchestrator import Orchestrator

orchestrator = Orchestrator(auto_configure=True)
orchestrator.configure_anthropic(api_key="sk-...")

# Ask for code generation - automatically routes to Claude
response = await orchestrator.execute_task(
    task="Write a Fibonacci function",
    capabilities=["code_generation"],
)

print(response.tool_used)    # claude_code_generation
print(response.provider)     # anthropic