olalonde/mcp-human

Overview

A Model Context Protocol server that enables AI assistants to request human input during tasks. It provides tools for the AI to ask questions, get approvals, or receive guidance from a human operator. Built in JavaScript, it integrates with MCP-compatible clients.

Best for

Best for
Developers building responsible AI agents that need supervised decision-making

Use cases

• Adding human verification steps to automated workflows
• Allowing AI agents to request clarification on ambiguous requests
• Implementing human-in-the-loop for sensitive actions in AI deployments

How to use

Install

claude mcp add human -- npx -y mcp-human@latest

Tools exposed

• MTURK_SANDBOX
• AWS_REGION
• AWS_PROFILE
• DEFAULT_REWARD
• FORM_URL

Tested with

Claude Desktop, Claude Code

Notes

A Model Context Protocol server that enables AI assistants to request human input during tasks. It provides tools for the AI to ask questions, get approvals, or receive guidance from a human operator. Built in JavaScript, it integrates with MCP-compatible clients.

24 stars on GitHub. Last updated 2025-05-08.

Use cases

• Adding human verification steps to automated workflows
• Allowing AI agents to request clarification on ambiguous requests
• Implementing human-in-the-loop for sensitive actions in AI deployments

Pros

• Simple integration with MCP ecosystems
• Enables safe delegation by adding human oversight
• Lightweight JavaScript implementation

Cons

• Requires human availability, limiting full automation
• Small community and limited documentation (24 stars)
• Potential latency from human response time

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.