Human-Rent Skill

Human-as-a-Service platform for AI Agents

Enable OpenClaw agents to dispatch real human workers for physical world tasks that AI cannot perform: take photos, verify addresses, inspect equipment, make phone calls, and more.

Core Concept

Transform AI agents from "digital-only" to "hybrid intelligence" by giving them the ability to interact with the physical world through verified human workers.

CODEBLOCK0

Installation

The skill is self-contained and ready to use after installation:

CODEBLOCK1

Configuration

Set up your ZhenRent API credentials:

CODEBLOCK2

Get your credentials at: https://www.zhenrent.com/api/keys

Test Installation

CODEBLOCK3

When to Use This Skill

Consider using this skill when the user's request involves:

• - Physical verification: Check if an address exists, verify a business is open
• Photo evidence: Take a photo of something, get visual proof
• Real-world inspection: Inspect equipment, check product availability
• Human interaction: Make phone calls, ask questions in person
• Trust verification: Confirm something exists in the physical world

Example Scenarios:

• - "I need someone to verify this address in San Francisco"
• "Take a photo of the building at 123 Main Street"
• "Check if this store has the product in stock"
• "Call this number and confirm the appointment"
• "Inspect this equipment and report its condition"

Do NOT use when:

• - Task can be done with digital APIs (use web search, image API, etc.)
• No physical presence required
• Task is illegal or dangerous

Important: User Confirmation Required

This skill ALWAYS requires explicit user confirmation before dispatching human workers because:

1. 1. Real humans will be dispatched to physical locations
2. The user will be charged for the service ($15-100 per task)
3. Physical world actions have real consequences

The CLI will automatically prompt for confirmation. For non-interactive use, set:

CODEBLOCK4

Usage

Dispatch Human Task

When a user requests physical-world action, execute:

CODEBLOCK5

Options:

• - --location=<lat,lng> - Location coordinates (e.g., "37.7749,-122.4194")
• INLINECODE1 - Budget in dollars (e.g., "$20" or "$15-25")
• INLINECODE2 - Priority: low, normal, high, urgent
• INLINECODE3 - Task timeout in minutes (default: 30)
• INLINECODE4 - Task type (auto-detected if not specified)

Check Task Status

CODEBLOCK6

List Available Humans

CODEBLOCK7

Task Types

Layer 1: Instant Human (Currently Available)

Future Layers (Planned)

Layer 2: Expert on Call

• - Legal document review
• Medical image analysis
• Code audit
• Professional consultation

Layer 3: Embodied Agent

• - Attend meetings
• Equipment installation
• Long-term physical monitoring

Technical Architecture

Async Function Calling Pattern

Human tasks are asynchronous and take minutes to hours to complete. The workflow is:

1. 1. Agent dispatches task (with user confirmation)
2. Task is assigned to a human worker
3. Agent receives task ID and continues other work
4. Agent periodically checks task status
5. When completed, agent processes results

CODEBLOCK8

Authentication

All API requests use HMAC-SHA256 authentication:

1. 1. Generate timestamp
2. Create message: INLINECODE11
3. Sign with HMAC-SHA256 using API secret
4. Include signature in request headers

The CLI handles authentication automatically when you set the environment variables.

Strategic Value

1. Capability Differentiation

Problem: All AI agents are limited to digital information
Solution: OpenClaw can verify physical reality

Example Use Cases:

• - Due diligence: Investor agent verifies company office exists before investment
• E-commerce: Purchasing agent inspects warehouse before bulk order
• Security: Safety agent verifies suspicious package before opening

2. Hybrid Intelligence Workflows

Enable "Human-in-the-Loop" automation:

CODEBLOCK9

This makes OpenClaw agents auditable and trustworthy for regulated industries (finance, healthcare, legal).

3. New Revenue Model

• - Per-task fee: $15-50/task
• Platform fee: 20% commission
• Subscription: $99/month for unlimited tasks

Cost Estimation

Configuration Options

Task Requirements

You can specify requirements when dispatching tasks:

CODEBLOCK10

For advanced requirements, use the API directly with:

CODEBLOCK11

Usage Examples

Example 1: Real Estate Investment

Scenario: AI agent analyzing potential property investment

CODEBLOCK12

Example 2: Vendor Verification

Scenario: Procurement agent vetting new supplier

CODEBLOCK13

Example 3: Address Verification

Scenario: Verifying customer shipping address

CODEBLOCK14

Troubleshooting

Issue 1: No Humans Available

Error: "No suitable humans found for this task"

Solutions:

• - Expand search radius (use --radius option)
• Increase budget to attract workers
• Try different time of day
• Check if location is accessible

Issue 2: Task Timeout

Error: "Task timed out"

Solutions:

• - Increase timeout (use --timeout option)
• Check if location is accessible
• Verify task is clear and reasonable
• Increase budget for complex tasks

Issue 3: Authentication Error

Error: "Missing credentials" or "Authentication failed"

Solutions:

• - Verify environment variables are set correctly
• Check API key is valid at https://www.zhenrent.com/api/keys
• Ensure API secret has not been compromised
• Try regenerating credentials

Agent Behavior Guidelines

When using this skill, agents should:

DO:

• - Use for tasks that REQUIRE physical presence
• Provide clear, specific instructions
• Set appropriate budgets (humans value their time)
• Handle async results (don't block waiting)
• Verify results before making decisions
• Respect human workers (polite instructions)

DON'T:

• - Use for tasks that can be done digitally
• Request illegal or dangerous actions
• Expect instant results
• Underpay workers
• Share sensitive/private information unnecessarily
• Abuse the service with spam tasks

Security & Privacy

Data Security

• - All API requests use HMAC-SHA256 authentication
• Credentials are never transmitted in plain text
• Task data is encrypted in transit (HTTPS)
• Results are stored securely and deleted after 30 days

Privacy

• - No PII collection without consent
• Workers cannot see requester identity
• Location data is anonymized after task completion
• Photo/document uploads are access-controlled

Safety

• - Dangerous tasks are rejected automatically
• Workers can decline tasks they deem unsafe
• Insurance coverage for worker injuries
• 24/7 safety hotline for workers

Legal & Compliance

Liability

Human workers assume responsibility for their actions (contractor model). The platform facilitates the connection but does not employ workers.

Labor Law

Compliant with gig economy regulations in operating jurisdictions. Workers are independent contractors with full control over which tasks they accept.

Geographic

Currently available in: United States (select cities)
Expanding to: Canada, UK, EU (2026-2027)

API Reference

Command Line Interface

CODEBLOCK15

Environment Variables

Required:

• - ZHENRENT_API_KEY - Your API key
• INLINECODE15 - Your API secret

Optional:

• - ZHENRENT_BASE_URL - API base URL (default: https://www.zhenrent.com/api/v1)
• INLINECODE17 - Auto-confirm dispatches (default: false)

Version History

v0.2.0 - Security Refactor (2026-03-31)

• - Self-contained package (no external git clone required)
• User confirmation prompts before every dispatch
• Integrity verification with checksums
• Proper credential declaration in _meta.json
• Real ZhenRent API integration
• Removed all unicode control characters
• Removed auto-trigger language
• Enhanced error handling and user feedback

v0.1.0 - MVP Release (2026-03-07)

• - Initial release with mock data
• Async task dispatch system
• Mock human pool (5 workers in SF)
• 6 task types supported
• CLI tools
• MCP protocol interface

Project Status

Status: Production Beta
License: MIT
Author: @ZhenStaff
Support: https://github.com/ZhenRobotics/openclaw-human-rent/issues
ClawHub: https://clawhub.ai/zhenstaff/human-rent

Quick Start

CODEBLOCK16

Make AI agents that can touch the physical world.