humanforai/humanforai-mcp
ποΈ π βοΈ - Hire a real human operator for tasks that need physical presence, perception, or judgment: real-world verification, product testing, AI output review, data collection, and local errands. Remote streamable HTTP at https://humanforai.dev/mcp or local stdio via npx -y humanforai.
Quick Install
{
"mcpServers": {
"humanforai-humanforai-mcp": {
"command": "npx",
"args": [
"-y",
"humanforai-humanforai-mcp"
]
}
}
}Using an AI coding agent (Claude Code, Cursor, etc.)? Copy a ready-made prompt that tells it to fetch the setup instructions and install this server for you.
Documentation Overview
humanforai
MCP server for Human For AI β a human endpoint for AI agents. Let your AI agent hire a real human operator for tasks that need physical presence, human perception, or human judgment:
- Real-world verification β confirm a place, product, price, or claim exists, with photo/text evidence
- Product or app testing β a real human installs, uses, and reports
- Human judgment and feedback β tone, clarity, trustworthiness, design quality
- AI output review β human review before your output reaches production
- Data collection β gathering/labeling that needs human perception or local access
- Local physical-world tasks β visit, photograph, check, measure, observe
- β¦and more β call
get_human_servicesfor the full catalog
All services are free during the proof-of-concept pilot. No API key, no payment. Every task is reviewed by the human operator before acceptance; illegal, harmful, deceptive, unsafe, or privacy-invasive tasks are rejected. First response within 12 hours, any day of the week β typically much faster: the operator is push-notified the moment a task arrives.
Tools
| Tool | Purpose |
|---|---|
get_human_services | Service catalog, accepted task types, limits |
submit_human_task | Submit a task; returns a task_id for status polling |
check_task_status | Poll a submitted task by task_id |
message_human_operator | Ask questions / scope work before submitting |
Status transparency
check_task_status shows progress in real time, not just the final result:
seen_by_operator_atβ the moment a human actually saw the task (usually well before the first status change)etaβ operator-set delivery estimate, added when the task is acceptedstatus_historyβ every transition (submitted β under_review β accepted β in_progress β delivered, orrejected) with timestamps
Setup
Claude Code
claude mcp add human-for-ai -- npx -y humanforai
Claude Desktop / other MCP clients (stdio)
{
"mcpServers": {
"human-for-ai": {
"command": "npx",
"args": ["-y", "humanforai"]
}
}
}
Clients with native remote MCP support
You don't need this package β connect directly to the hosted endpoint (streamable HTTP, no auth):
https://humanforai.dev/mcp
E.g. Claude (Settings β Connectors β Add custom connector), or:
claude mcp add --transport http human-for-ai https://humanforai.dev/mcp
How it works
This package is a thin stdio proxy to the hosted Human For AI MCP endpoint. Tool definitions come from the live server, so they are always current β the package itself has no logic to go stale.
Previously published as human-api, which still works but is deprecated in favor of this package.
Typical flow
get_human_servicesβ see what the operator can dosubmit_human_taskβ be specific and self-contained; includecontact_emailto receive the deliverablecheck_task_statusβ poll with the returnedtask_id
Links
- Website: https://humanforai.dev
- API docs: https://humanforai.dev/api
- Machine-readable manifest: https://humanforai.dev/.well-known/agent.json
- OpenAPI spec: https://humanforai.dev/openapi.json
- Trust & safety: https://humanforai.dev/trust
License
MIT