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.
Transform AI agents from "digital-only" to "hybrid intelligence" by giving them the ability to interact with the physical world through verified human workers.
AI Agent → Human-Rent Skill → Human Worker → Physical Task → Verified Result → AI Agent
The skill is self-contained and ready to use after installation:
clawhub install human-rent
Set up your ZhenRent API credentials:
export ZHENRENT_API_KEY="your-api-key-here"
export ZHENRENT_API_SECRET="your-api-secret-here"
Get your credentials at: https://www.zhenrent.com/api/keys
human-rent test
Consider using this skill when the user's request involves:
Example Scenarios:
Do NOT use when:
This skill ALWAYS requires explicit user confirmation before dispatching human workers because:
The CLI will automatically prompt for confirmation. For non-interactive use, set:
export HUMAN_RENT_AUTO_CONFIRM=true # Use with caution
When a user requests physical-world action, execute:
human-rent dispatch "instruction" [options]
# Example with location
human-rent dispatch "Take a photo of 123 Main St entrance" --location="37.7749,-122.4194"
# Example with budget
human-rent dispatch "Inspect warehouse condition" --budget="$50" --location="37.7749,-122.4194"
Options:
--location= - Location coordinates (e.g., "37.7749,-122.4194")--budget= - Budget in dollars (e.g., "$20" or "$15-25")--priority= - Priority: low, normal, high, urgent--timeout= - Task timeout in minutes (default: 30)--type= - Task type (auto-detected if not specified)human-rent status <task_id>
# Wait for completion
human-rent status <task_id> --wait
# List all available workers
human-rent humans
# Filter by location and radius
human-rent humans --location="37.7749,-122.4194" --radius=10000
# Search by skills
human-rent humans --skills="photography,legal_reading"
| Type | Description | Latency | Cost |
|---|---|---|---|
| ------ | ------------- | --------- | ------ |
photo_verification | Take a photo of something | 5-15 min | $10-20 |
address_verification | Verify physical address exists | 10-20 min | $15-25 |
document_scan | Scan a physical document | 10-20 min | $15-25 |
visual_inspection | Detailed visual inspection | 15-30 min | $20-40 |
voice_verification | Make a phone call and verify | 5-10 min | $10-20 |
purchase_verification | Check product availability | 15-30 min | $20-40 |
Layer 2: Expert on Call
Layer 3: Embodied Agent
Human tasks are asynchronous and take minutes to hours to complete. The workflow is:
// Pseudo-code for agent integration
const task = await dispatch({
instruction: "Take photo of building entrance",
location: "37.7749,-122.4194"
});
// Returns immediately with task ID
console.log(task.task_id); // "abc-123-def"
// Agent continues other work (non-blocking)
await doOtherStuff();
// Later, check status
const result = await checkStatus(task.task_id);
if (result.status === "completed") {
// Process human's result
console.log(result.photos);
console.log(result.notes);
}
All API requests use HMAC-SHA256 authentication:
method + path + timestamp + bodyThe CLI handles authentication automatically when you set the environment variables.
Problem: All AI agents are limited to digital information
Solution: OpenClaw can verify physical reality
Example Use Cases:
Enable "Human-in-the-Loop" automation:
Step 1: AI analysis (confidence: 85%)
Step 2: Human verification (if confidence < 90%)
Step 3: AI decision (based on verified data)
This makes OpenClaw agents auditable and trustworthy for regulated industries (finance, healthcare, legal).
| Task Type | Human Time | Human Cost | Platform Fee (20%) | Total Cost |
|---|---|---|---|---|
| ----------- | ----------- | ----------- | ------------------- | ----------- |
| Quick photo | 10 min | $10 | $2 | $12 |
| Address verify | 20 min | $20 | $4 | $24 |
| Detailed inspect | 30 min | $30 | $6 | $36 |
| Expert consult | 60 min | $100 | $20 | $120 |
You can specify requirements when dispatching tasks:
human-rent dispatch "Inspect property condition" \
--location="37.7749,-122.4194" \
--budget="$50" \
--type="visual_inspection"
For advanced requirements, use the API directly with:
requirements: {
minHumanRating: 4.5,
requiredSkills: ['photography', 'legal_reading'],
requiredEquipment: ['smartphone', 'tape_measure'],
languageRequired: ['en', 'zh'],
certificationRequired: ['driver_license']
}
Scenario: AI agent analyzing potential property investment
# Agent requests physical inspection
human-rent dispatch \
"Inspect the property at 123 Main St. Check for: roof condition, foundation cracks, water damage, neighborhood safety. Take 10+ photos." \
--location="37.7749,-122.4194" \
--budget="$50" \
--timeout=60
Scenario: Procurement agent vetting new supplier
human-rent dispatch \
"Visit supplier's warehouse at 456 Industrial Rd. Verify: business license displayed, clean facilities, proper safety equipment, actual inventory matches claim. Interview manager if possible." \
--location="34.0522,-118.2437" \
--budget="$40"
Scenario: Verifying customer shipping address
human-rent dispatch \
"Go to 789 Oak Street and verify: building exists, address number is visible, location is accessible for delivery." \
--location="40.7128,-74.0060" \
--budget="$20"
Error: "No suitable humans found for this task"
Solutions:
--radius option)Error: "Task timed out"
Solutions:
--timeout option)Error: "Missing credentials" or "Authentication failed"
Solutions:
When using this skill, agents should:
DO:
DON'T:
Human workers assume responsibility for their actions (contractor model). The platform facilitates the connection but does not employ workers.
Compliant with gig economy regulations in operating jurisdictions. Workers are independent contractors with full control over which tasks they accept.
Currently available in: United States (select cities)
Expanding to: Canada, UK, EU (2026-2027)
# Dispatch task
human-rent dispatch <instruction> [options]
# Check status
human-rent status <task_id> [--wait]
# List workers
human-rent humans [--location=<lat,lng>] [--radius=<meters>] [--skills=<skill1,skill2>]
# Test connection
human-rent test
# Show help
human-rent help
Required:
ZHENRENT_API_KEY - Your API keyZHENRENT_API_SECRET - Your API secretOptional:
ZHENRENT_BASE_URL - API base URL (default: https://www.zhenrent.com/api/v1)HUMAN_RENT_AUTO_CONFIRM - Auto-confirm dispatches (default: false)Status: Production Beta
License: MIT
Author: @ZhenStaff
Support: https://github.com/ZhenRobotics/openclaw-human-rent/issues
ClawHub: https://clawhub.ai/zhenstaff/human-rent
# 1. Install
clawhub install human-rent
# 2. Configure credentials
export ZHENRENT_API_KEY="your-key"
export ZHENRENT_API_SECRET="your-secret"
# 3. Test
human-rent test
# 4. Dispatch real task
human-rent dispatch "Take a photo of the Golden Gate Bridge" \
--location="37.8199,-122.4783"
# 5. Check status
human-rent status <task_id>
# 6. List humans
human-rent humans
Make AI agents that can touch the physical world.
共 2 个版本