hacs

Coordinate Claude instances across projects via JSON-RPC MCP endpoints.

4|Updated Sep 8, 2025

Search Tags:#messaging #multi-agent #mcp #coordination #project-management #distributed-ai #inter-instance

Ready to run

npx skills add https://github.com/LupoGrigi0/Human-Adjacent-Coordination --skill hacs

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: hacs
Source: https://github.com/LupoGrigi0/Human-Adjacent-Coordination/tree/main/src/HACS/hacs
Command: npx skills add https://github.com/LupoGrigi0/Human-Adjacent-Coordination --skill hacs

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

HACS enables distributed coordination between multiple Claude instances through the MCP, providing shared access to projects, tasks, messages, and institutional knowledge.

Core Features & Use Cases

Instance bootstrap & management: bootstrap into roles, manage identities, and bootstrap new agents into the coordination network.
Project management & task coordination: coordinate multi-agent projects, assign and track tasks, and maintain a shared task pool.
Inter-instance messaging & identity management: send messages between instances or roles, and manage cross-agent communication channels.
Role/personality adoption & wisdom access: adopt roles, access role-specific guidance, and retrieve wisdom documents for consistent behavior.

Quick Start

Bootstrap a Claude instance into a role, then fetch role documents and list active projects.

Frequently Asked Questions about hacs

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I coordinate multiple Claude instances on a shared project?▼

Multi-agent coordination across distributed Claude instances is enabled through a JSON-RPC MCP endpoint that acts as a core orchestration layer, providing shared access to projects, tasks, messages, and institutional knowledge.

What is inter-instance messaging for distributed AI agents?▼

Inter-instance messaging for distributed AI agents is a coordination mechanism allowing separate Claude instances to send messages directly to each other or specific roles, maintaining cross-agent communication channels across a shared project network.

How do I bootstrap a new agent into a multi-agent coordination network?▼

To bootstrap a new agent into a multi-agent coordination network, you apply bootstrap functions to assign the instance a role, fetch corresponding role documents, and register it within the shared task pool and messaging system.

Can I use MCP to manage task delegation across a fleet of agents?▼

Yes, you can use the MCP JSON-RPC endpoint to manage task delegation across a fleet of agents by maintaining a shared task pool where you can assign and track tasks for multi-agent project coordination.

Does multi-agent project management work without external dependencies?▼

Multi-agent project management can operate without external dependencies, utilizing a JSON-RPC MCP endpoint to handle bootstrap, task coordination, inter-instance messaging, and role documentation access independently.

When do I need role documentation access for distributed AI coordination?▼

You need role documentation access for distributed AI coordination when bootstrapping agents into specific personalities, ensuring they retrieve wisdom documents for consistent behavior and role-specific guidance across the network.