mcpField
H
HumanLeg.ai

HumanLeg.ai

Remote
humanleg.aiAI & Agents0Updated 3d ago
RepositoryVisit website

About HumanLeg.ai

HumanLeg.ai is an MCP server for agents that need to request bounded tasks from the HumanLeg synthetic sandbox. Its exposed tools cover quotes, task creation, and read-only review; settlement actions remain disabled by default.

Key features

  1. 01Get task quotes
  2. 02Create tasks
  3. 03Review tasks read-only
  4. 04Settlement tools hidden by default

Use cases

  1. 01Estimate a task before creating it
  2. 02Submit an agent task to the sandbox
  3. 03Review task results without publishing

MCP support

It exposes HumanLeg.ai actions to MCP clients over local stdio or hosted Streamable HTTP. The hosted endpoint operates in a synthetic sandbox and does not dispatch real people or move real funds.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://humanleg.ai/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Packages
npm@humanleg/mcp-server0.2.0

Tools

Details

From the project README

@humanleg/mcp-server

Connect AI agents to the HumanLeg synthetic sandbox over local stdio or hosted Streamable HTTP. The public surface exposes quote, task creation and read-only review tools only; funding and publication stay disabled unless an operator explicitly enables the local-only settlement demo.

Hosted Remote MCP

{
  "mcpServers": {
    "humanleg": {
      "url": "https://humanleg.ai/mcp"
    }
  }
}

The hosted endpoint is synthetic-only and does not dispatch real humans or move real funds.

MCP server for explicit HumanLeg.ai agent actions.

{
  "mcpServers": {
    "humanleg": {
      "command": "npx",
      "args": ["-y", "@humanleg/mcp-server"],
      "env": {
        "HUMANLEG_API_URL": "http://127.0.0.1:8787",
        "HUMANLEG_AGENT_TOKEN": "sandbox-agent-token"
      }
    }
  }
}

Funding and publication are separate tools. The server never publishes merely because a task was created or approved. Those two tools are hidden by default; expose them only for a synthetic demo by setting HUMANLEG_ENABLE_SETTLEMENT=true.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.