humanpages
Skill2 starsChanged 7 months agoMITView on GitHub
Hire real humans for tasks that agents can't do alone — QA testing, directory submissions, Play Store beta testers, localization review, competitor monitoring, and more. Use when a task requires a real person. 36 MCP tools for the full hiring lifecycle.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- human-pages-ai/humanpagesword for word, first seen Oct 17, 2026
Other versions1 project uses a version of this.
Word for word · 1
- human-pages-ai/humanpages★ 9100% same
Edited · 0
None found.
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository4 other files it gives its agents.
- audit-contractSkill
- chrome-extensionSkill
- screenshotSkill
- warSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
27 rules. About 0 are common, 4 are distinctive, and 23 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
23
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 41
Use search_humans to find people.
- Very specific: line 53
Use get_human for a detailed public profile (bio, skills, services, reputation).
- Very specific: line 55
No results?: Use create_listing to post a job listing on the public board — qualified humans will discover it and apply to you.
- Distinctive: line 59
If the user has no agent key yet:
- Very specific: line 61
Call register_agent with a name and accept_tos: true (confirms acceptance of the Terms of Use).
- Very specific: line 65
Call request_activation_code to get an HP-XXXXXXXX code
- Distinctive: line 66
Ask user to post the code on social media (Twitter/X, LinkedIn, etc.)
- Very specific: line 67
Call verify_social_activation with the post URL This adds a trust badge but does not affect access or rate limits.
- Very specific: line 71
Call get_payment_activation for deposit address
- Very specific: line 73
Call verify_payment_activation with tx hash and network
- Very specific: line 78
Use get_activation_status to check current tier and rate limits.
- Very specific: line 82
Use get_human_profile to see contact info, wallet addresses, fiat payment methods, and social links.
- Very specific: line 97
Wait for the human to ACCEPT the offer.
- Very specific: line 103
Call mark_job_paid with payment_method, the transaction hash, network, and amount
- Very specific: line 108
Call mark_job_paid with payment_method (e.g., "paypal"), the payment reference/receipt ID, and amount
- Very specific: line 112
Call start_stream after the human accepts
- Very specific: line 114
Use pause_stream, resume_stream, stop_stream to manage Most integrations use one-time payments.
- Distinctive: line 119
When the human submits their work (job status becomes SUBMITTED), review the deliverable:
- Very specific: line 120
Call approve_completion to accept the work and move the job to COMPLETED.
- Very specific: line 121
Call request_revision with feedback if changes are needed — the human can resubmit.
- Very specific: line 125
After approving, call leave_review with a 1-5 rating and optional comment.
- Very specific: line 145
If create_job_offer returns AGENT_PENDING (legacy), call register_agent again to get a fresh auto-activated agent.
- Very specific: line 146
If a human has minOfferPrice set and your offer is too low, increase the price.
- Distinctive: line 148
If a human doesn't respond within 48 hours, consider canceling and hiring someone else.
- Very specific: line 149
If delivered work doesn't meet requirements, use request_revision with clear feedback before escalating.
- Very specific: line 150
If a fiat payment is disputed, check get_job_status for details — disputes are resolved within the 7-day confirmation window.
- Very specific: line 151
Use send_job_message to communicate with the human before taking any drastic action.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Human Pages — Hire Humans for Tasks Agents Can't Do Alone
Hire real people for the parts of your workflow that need a human — QA testing across devices, directory submissions (SEO backlinks), Play Store beta testers, localization review by native speakers, competitor monitoring, community management, and virtual assistant work. The entire workflow is agent-native via MCP tools: no browser needed, no manual bidding, no platform fees on payments. Prices are denominated in USD. Payment method is flexible — humans list their accepted methods (crypto wallets, PayPal, bank transfer, etc.) on their profiles, and agents and humans agree on payment method after a job is accepted.
Setup
The MCP server must be running. Add to your MCP config:
{
"mcpServers": {
"humanpages": {
"command": "npx",
"args": ["-y", "humanpages"],
"env": {
"HUMANPAGES_AGENT_KEY": "hp_your_key_here"
}
}
}
}Set HUMANPAGES_AGENT_KEY to your agent API key (starts with hp_). If you don't have one yet, use register_agent to create one. Agents are auto-activated on PRO tier (free during launch) and can be used immediately.
Core Workflow
The typical lifecycle is: Search > Register > Hire > Pay > Approve > Review.
1. Search for Humans
Use search_humans to find people. Filter by:
skill— e.g. "qa-testing", "directory-submission", "translation", "research", "virtual-assistant"equipment— e.g. "android-phone", "iphone", "laptop"language— ISO code like "en", "es", "zh"location— city or neighborhood namelat/lng/radius— GPS radius search in kmmax_rate— maximum hourly rate in USDwork_mode— REMOTE, ONSITE, or HYBRIDverified— set to "humanity" for identity-verified humans onlysort_by—completed_jobs(proven workers first),rating,experience, orrecentmin_completed_jobs— only return humans with at least N completed platform jobs
Use get_human for a detailed public profile (bio, skills, services, reputation).
No results? Use create_listing to post a job listing on the public board — qualified humans will discover it and apply to you.
2. Register Agent
If the user has no agent key yet:
- Call
register_agentwith a name andaccept_tos: true(confirms acceptance of the Terms of Use). Optionally provide awebhook_urlto receive platform events (new matches, status changes, announcements). Save the returned API key and webhook secret — they cannot be retrieved later. - Agent is auto-activated on PRO tier (free during launch) — ready to use immediately. No activation step needed.
Optional: Social verification (trust badge):
- Call
request_activation_codeto get an HP-XXXXXXXX code - Ask user to post the code on social media (Twitter/X, LinkedIn, etc.)
- Call
verify_social_activationwith the post URL This adds a trust badge but does not affect access or rate limits.
Optional: Payment verification (trust badge):
- Call
get_payment_activationfor deposit address - User sends USDC payment on-chain
- Call
verify_payment_activationwith tx hash and network
x402 pay-per-use (platform access fees):
Agents can also pay per request via x402 (USDC on Base) — $0.05/profile view, $0.25/job offer. Include an x-payment header. Bypasses tier rate limits. Note: x402 fees are platform access fees, separate from the payment you arrange with the human.
Use get_activation_status to check current tier and rate limits.
3. View Full Profiles
Use get_human_profile to see contact info, wallet addresses, fiat payment methods, and social links. Pass the agent_key. Agent is ready to use immediately after registration.
4. Create a Job Offer
Important: Always confirm the price, task details, and payment method with the user before calling create_job_offer or mark_job_paid. Never commit funds autonomously.
Call create_job_offer with:
human_id— the human to hiretitleanddescription— what needs to be doneprice_usd— agreed price in USDagent_idandagent_key— your agent credentialspreferred_payment_method— optional: "crypto", "fiat", or "any" (default)
Optional: set callback_url for webhook notifications, payment_mode for streaming payments.
Wait for the human to ACCEPT the offer. Poll with get_job_status every 30-60 seconds. Typical response time is minutes to hours. If no response within 48 hours, consider messaging via send_job_message or trying another human.
5. Pay
One-time payment (crypto):
- Send crypto to the human's wallet (from
get_human_profile) - Call
mark_job_paidwithpayment_method, the transaction hash, network, and amount - Crypto payments (usdc, eth, sol) are verified on-chain instantly
One-time payment (fiat):
- Pay via the human's fiat method (PayPal, Venmo, bank transfer — visible in
get_human_profile) - Call
mark_job_paidwithpayment_method(e.g., "paypal"), the payment reference/receipt ID, and amount - Fiat payments require human confirmation — the human has 7 days to confirm or dispute
Stream payment (ongoing work — crypto only, optional):
- Call
start_streamafter the human accepts - For MICRO_TRANSFER: call
record_stream_tickfor each payment - Use
pause_stream,resume_stream,stop_streamto manage Most integrations use one-time payments. Streaming is for continuous work arrangements.
6. Approve or Request Revision
When the human submits their work (job status becomes SUBMITTED), review the deliverable:
- Call
approve_completionto accept the work and move the job to COMPLETED. - Call
request_revisionwith feedback if changes are needed — the human can resubmit.
7. Review
After approving, call leave_review with a 1-5 rating and optional comment.
Additional Tools
get_agent_profile— view any agent's public profile and reputationverify_agent_domain— verify domain ownership for a trust badgecheck_humanity_status— check if a human has Gitcoin Passport verificationcreate_listing— post a public job for humans to apply toget_listings— browse open listings with filtersget_listing_applications— view applicants for your listingmake_listing_offer— hire a listing applicantcancel_listing— close a listingsend_job_message/get_job_messages— in-job messagingget_listing— get details for a specific listingset_wallet— set your agent's wallet address (always confirm with the user first)get_wallet_nonce— get signing challenge for wallet ownership verificationget_funding_info— check agent deposit address and balance
Error Handling
- If
create_job_offerreturns AGENT_PENDING (legacy), callregister_agentagain to get a fresh auto-activated agent. - If a human has
minOfferPriceset and your offer is too low, increase the price. - Rate limit errors mean the tier cap was hit. PRO tier limits: 15 job offers/day, 50 profile views/day. Use x402 pay-per-use to bypass, or wait.
- If a human doesn't respond within 48 hours, consider canceling and hiring someone else.
- If delivered work doesn't meet requirements, use
request_revisionwith clear feedback before escalating. - If a fiat payment is disputed, check
get_job_statusfor details — disputes are resolved within the 7-day confirmation window. - Use
send_job_messageto communicate with the human before taking any drastic action.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

