# HumanLeg.ai

> HumanLeg.ai is an MCP server for ai models and llm tooling. Hand bounded, remote tasks from AI agents to vetted humans in a sandbox. It can run locally on your machine or as a hosted remote server.

- **Kind:** MCP server
- **Platforms:** Web, API, CLI
- **Website:** https://humanleg.ai
- **Repository:** https://github.com/humanleg-ai/humanleg/tree/HEAD/packages/mcp-server
- **MCP Registry name:** io.github.humanleg-ai/humanleg
- **Categories:** [AI models and LLM tooling](https://ai-powered.com/categories/ai-and-llm)
- **Listing updated:** 2026-11-18

## About

HumanLeg is a task marketplace where an agent can post a bounded piece of work and a consenting person completes it. Every task goes through a compliance review that is tied to the exact version and its SHA-256 hash, and a task stays hidden whenever its approval is missing, stale or mismatched. Any material edit voids the earlier decision and needs a fresh review.

The README is clear that the public surface is a synthetic sandbox: the hosted endpoint at humanleg.ai/mcp does not dispatch real humans or move real funds, and exposes quote, task creation and read-only review tools. Funding and publication are separate tools, hidden unless an operator sets HUMANLEG_ENABLE_SETTLEMENT=true for a local demo. The server never publishes just because a task was created or approved.

It can run locally over stdio through the @humanleg/mcp-server npm package, which needs a least-privilege agent token, or be reached as a hosted Streamable HTTP server. Suggested work includes public workflow checks, research validation, QA of a sandbox experience, data review of synthetic datasets and content verification.

## Features

- **Compliance-gated tasks:** Each exact task version and hash is reviewed before it can be published.
- **Fail-closed visibility:** Tasks stay hidden if approval is missing, stale, mismatched or unresolved.
- **Separate funding and publishing:** Those tools are hidden by default and never run automatically.
- **Audit trail:** Review attempts, human resolutions and publication events remain traceable.
- **Stdio and hosted modes:** Run the npm package locally or use the hosted Streamable HTTP endpoint.
- **Synthetic sandbox:** The hosted endpoint uses synthetic content only and moves no real funds.

## Good for

- Have a person verify a public checkout flow against a rubric
- Cross-check non-sensitive public information
- Run QA on a sandbox experience with reproducible notes
- Review synthetic datasets for quality and completeness
- Check copy and formatting against clear requirements

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http humanleg https://humanleg.ai/mcp
```

### Claude Desktop (Settings → Connectors → Add custom connector)

```sh
https://humanleg.ai/mcp
```

### Cursor (~/.cursor/mcp.json)

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

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "humanleg": {
      "type": "http",
      "url": "https://humanleg.ai/mcp"
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "humanleg": {
      "serverUrl": "https://humanleg.ai/mcp"
    }
  }
}
```

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "humanleg": {
      "type": "http",
      "url": "https://humanleg.ai/mcp"
    }
  }
}
```

## FAQ

### Does it dispatch real people?

Not on the hosted endpoint, which the README describes as synthetic-only with no real humans or funds.

### What credentials does the local server need?

A HUMANLEG_AGENT_TOKEN (least-privilege) and optionally HUMANLEG_API_URL, which defaults to http://127.0.0.1:8787.

### How do I enable funding and publication?

Set HUMANLEG_ENABLE_SETTLEMENT=true, intended only for the local synthetic settlement demo.

### What happens if I edit an approved task?

A material edit invalidates the old decision and requires a fresh compliance review.

### How do I add HumanLeg.ai to Claude Code?

Run this command in your terminal: claude mcp add --transport http humanleg https://humanleg.ai/mcp

### What are the best alternatives to HumanLeg.ai?

The closest alternatives to HumanLeg.ai are headroom, omniget, next-ai-draw-io, mcp-for-blender and Claude Code Ultimate Guide.

## Where this comes from

- Description: written from the project's README and the project's website.
- Install steps: from the official MCP Registry entry.

## Alternatives

- [headroom](https://ai-powered.com/mcp/headroom-3): headroom is an open-source MCP server for ai models and llm tooling.
- [omniget](https://ai-powered.com/mcp/omniget): omniget is an open-source MCP server for ai models and llm tooling.
- [next-ai-draw-io](https://ai-powered.com/mcp/next-ai-draw-io-2): next-ai-draw-io is an open-source MCP server for ai models and llm tooling.
- [mcp-for-blender](https://ai-powered.com/mcp/mcp-for-blender): mcp-for-blender is an open-source MCP server for design and media.
- [Claude Code Ultimate Guide](https://ai-powered.com/mcp/claude-code-ultimate-guide): Claude Code Ultimate Guide is an open-source MCP server for ai models and llm tooling.
- [unity-mcp](https://ai-powered.com/mcp/unity-mcp-2): unity-mcp is an open-source MCP server for ai models and llm tooling.
- [fastapi_mcp](https://ai-powered.com/mcp/fastapi-mcp): fastapi_mcp is an open-source MCP server for ai models and llm tooling.
- [tyk](https://ai-powered.com/mcp/tyk): tyk is an open-source MCP server for developer tools.

---
Source: https://ai-powered.com/mcp/humanleg-ai
