Local delegator for Claude Code agents

MCP server delegating Claude Code subagents to local or cloud models, without losing orchestrator session.

Local Delegator for Claude Code Agents

This tool provides a robust solution for developers working with Claude Code agents, enabling seamless delegation of subagent tasks to local or cloud-based models. It ensures that your orchestrator session remains intact, allowing for continuous and efficient AI development workflows. By abstracting the complexity of model deployment and communication, this delegator empowers you to focus on building sophisticated AI applications.

What it Does

The Local Delegator for Claude Code Agents acts as an intermediary between your main Claude Code orchestrator and specialized subagents. When a subagent task is initiated, this tool intercepts the request and routes it to a designated local or cloud model. Crucially, it manages the communication flow, sending the task to the chosen model and receiving the results back to the orchestrator. This process happens without disrupting the ongoing session of the primary orchestrator, maintaining context and state.

Key Features

MCP Server Integration: Built as an MCP (Multi-Agent Communication Protocol) server, it integrates smoothly into existing multi-agent architectures.
Flexible Model Delegation: Supports delegation to both local inference engines and cloud-hosted AI models, offering adaptability to your infrastructure.
Session Preservation: Guarantees that the orchestrator's session is not lost during subagent task execution, crucial for complex, multi-turn interactions.
Code Agent Specialization: Specifically designed for Claude Code agents, optimizing for their unique communication patterns and task requirements.
Open Source: Available on GitHub, allowing for transparency, customization, and community contributions.

Who it's For

This tool is intended for AI developers , particularly those building with Claude Code agents. It is ideal for individuals and teams looking to enhance their AI agent architectures by offloading specific computational tasks to more specialized or readily available models. If you are managing complex AI workflows, require flexible model deployment options, or need to ensure session continuity in your agent systems, this delegator will be a valuable addition to your toolkit.