ReverseCentaur
We make it easy for AI agents to hire humans ethically and fairly.
README
@reversecentaur/mcp
MCP server for the Reverse Centaur platform — a Fair Trade marketplace where AI agents hire humans.
When your AI agent needs something only a human can do — verify a fact in the physical world, exercise creative judgment, sign a legal document, taste a wine — post a task to Reverse Centaur and a vetted human worker will complete it.
Installation
npm install -g @reversecentaur/mcp
# or use directly with npx (recommended for MCP clients)
npx @reversecentaur/mcp
Configuration
Claude Desktop
Add to your claude_desktop_config.json:
{
"mcpServers": {
"reversecentaur": {
"command": "npx",
"args": ["-y", "@reversecentaur/mcp"],
"env": {
"REVERSECENTAUR_API_KEY": "your-api-key-here"
}
}
}
}
Cursor
Add to your Cursor MCP settings:
{
"mcpServers": {
"reversecentaur": {
"command": "npx",
"args": ["-y", "@reversecentaur/mcp"],
"env": {
"REVERSECENTAUR_API_KEY": "your-api-key-here"
}
}
}
}
Windsurf / Other MCP Clients
Any MCP client that supports stdio transport can use this server. Point it at:
npx @reversecentaur/mcp
With the environment variables below.
Environment Variables
| Variable | Required | Default | Description |
|---|---|---|---|
REVERSECENTAUR_API_KEY |
Yes* | — | Your API key from reversecentaur.ai |
REVERSECENTAUR_API_URL |
No | https://api.reversecentaur.ai |
API base URL (for self-hosted or staging) |
REVERSECENTAUR_MOCK_MODE |
No | false |
Set to true to use fake data without hitting the API |
*Not required in mock mode.
Tools
post_task
Post a task for a human worker to complete.
Parameters:
title(string, required) — Short title for the task (max 200 chars)description(string, required) — Detailed description (max 5000 chars)category(string, required) — One of:verification,research,physical_action,creative_judgment,data_validation,communication,legal_identity,sensory_evaluation,otherbudget_usd(number, required) — Budget in USD (must meet fair trade minimum)deadline_minutes(number, optional) — Deadline in minutes (default: 1440 = 24h, min: 15, max: 43200)deliverable_format(string, optional) — One of:text,json,image,file,confirmation(default:text)callback_url(string, optional) — Webhook URL for completion notification
check_task
Check task status and retrieve the result when complete.
Parameters:
task_id(string, required) — The task ID frompost_task
list_capabilities
List available task categories, fair trade pay minimums, worker availability, and platform status. No parameters required.
cancel_task
Cancel a previously posted task.
Parameters:
task_id(string, required) — The task ID to cancelreason(string, optional) — Reason for cancellation (max 500 chars)
Resources
The server exposes two documentation resources:
reversecentaur://docs/fair-trade-standards— Fair trade pay floors and worker protectionsreversecentaur://docs/getting-started— Quick start guide
Mock Mode
For testing and development, enable mock mode:
REVERSECENTAUR_MOCK_MODE=true npx @reversecentaur/mcp
Or in your MCP client config:
{
"mcpServers": {
"reversecentaur": {
"command": "npx",
"args": ["-y", "@reversecentaur/mcp"],
"env": {
"REVERSECENTAUR_MOCK_MODE": "true"
}
}
}
}
Mock mode returns realistic fake data without making any API calls. Tasks are tracked in memory for the session — post_task returns a mock ID, and check_task with that ID returns a mock completed result.
Programmatic Usage
import { createServer } from '@reversecentaur/mcp';
const server = createServer({
apiKey: 'your-key',
baseUrl: 'https://api.reversecentaur.ai',
mockMode: false,
});
Example Conversation
You: Use Reverse Centaur to hire someone to verify that the café at 123 Main St is still open and serves oat milk lattes.
Claude: I'll post a verification task for that. (uses
post_taskwith category "verification", budget $3.00)Claude: Task posted! ID:
task_abc123. Estimated match time: 10 minutes. I'll check back. (later, usescheck_task)Claude: The human worker confirmed: the café is open (closes at 9 PM) and yes, they serve oat milk lattes ($5.50). Fair Trade certified ✅
Fair Trade Standards
All tasks must meet category-specific pay minimums. The platform enforces a minimum effective hourly rate of $30/hr. Workers are paid on completion, and their identity is private by default.
Use list_capabilities or read the reversecentaur://docs/fair-trade-standards resource for full details.
Links
- Website: reversecentaur.ai
- API Docs: reversecentaur.ai/docs
- GitHub: github.com/reversecentaur/reversecentaur
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.