What is the MCP Agent Orchestrator?

It is a specialized skill for Claude Code that provides standardized patterns for coordinating multiple AI agents to solve complex tasks using the Model Context Protocol (MCP).

What is the benefit of the Evaluator-Optimizer pattern?

This pattern creates a feedback loop where a generator agent produces work and an evaluator agent critiques it, ensuring the final output meets specific quality thresholds.

Does it support OpenAI Swarm patterns?

Yes, it includes a Swarm-compatible implementation that allows agents to hand off tasks to one another seamlessly while maintaining state.

Can I use this for long-duration research?

Absolutely. The Deep Orchestrator pattern is specifically designed for long-horizon research tasks involving multi-step reasoning and knowledge extraction.

What does GF(3) category ZERO mean for this skill?

In this context, GF(3) ZERO indicates a coordination role, where the skill acts as a neutral hub or planner to balance generative (plus) and evaluative (minus) agent roles.

MCP Agent Orchestrator

Name: MCP Agent Orchestrator
Author: plurigrid

byplurigrid

•

Productivity & Workflow

Orchestrates complex multi-agent workflows using industry-standard patterns like swarms, planners, and evaluator-optimizers.

The MCP Agent Orchestrator skill empowers Claude to manage sophisticated multi-agent architectures using the Model Context Protocol. It provides a suite of coordination patterns including OpenAI-style Swarms with seamless handoffs, Planner-Worker configurations for parallel task execution, and Evaluator-Optimizer loops for iterative output refinement. By leveraging these composable workflows, developers can build robust autonomous systems that handle long-horizon research, intent-based routing, and specialized domain tasks with high reliability and GF(3) balanced coordination.

Key Features

01Deep Orchestration for long-horizon research and reasoning

02Iterative Evaluator-Optimizer loops for self-correcting output

038 GitHub stars

04Swarm-compatible agent handoffs and state management

05Intent-based routing and embedding-based classification

06Planner-Worker orchestration for parallel task decomposition

Use Cases

01Creating high-fidelity content through iterative generation and automated proofreading

02Building a multi-agent research team to synthesize complex technical data

03Automating support workflows with intelligent handoffs between specialized sub-agents

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add plurigrid/asi mcp-orchestrator

For use in Claude.ai and ChatGPT

Download Skill