Skip to content

Human Pages MCP server

io.github.human-pages-ai/humanpageshuman-pages-aistreamable-httpv1.4.6Scanned

Hire real humans for tasks agents can't do alone. 36 tools for the full hiring lifecycle.

40 tools: 1 can delete or cancel things, 3 do something that can't be undone, 20 change things and 16 only read.

Tools: 1 delete, 3 irreversible, 20 write, 16 read.
  • Reads private data · 2
  • Reads text others write · 5
  • Can send data out · 5
  • 1 finding

Tools listed without sign-in. Remote posture

Anlyon alternative

24 of these tools change something. Run those as governed actions.

Declare what an action changes, and Anlyon puts it under a shared limit, binds every approval to the exact request, and grades the outcome. It works on any HTTPS API you can describe as an action.

Anlyon Radar does not run, proxy or govern this MCP server. A tool your own code calls directly is not routed through Anlyon and is not governed by it.

Import an OpenAPI spec →

What the Human Pages MCP server can do

The Human Pages MCP server is published by the GitHub account human-pages-ai. It has 40 tools. It can delete or cancel things (1 tool: cancel_listing), do things that cannot be undone, such as sending, paying or publishing (create_job_offer, verify_payment_activation and send_job_message) and create or change data (20 tools: register_agent, get_wallet_nonce, set_wallet, verify_agent_domain and 16 more). 16 tools only read.

2 tools read private data (get_human_profile and get_job_messages); 5 tools bring in text other people can write (get_human, get_human_profile, get_job_messages and 2 more); 5 tools can send data out of your system (register_agent, create_job_offer, send_job_message and 2 more). On its own it has all three parts of a prompt-injection exfiltration path: private data, text others write, and a way out.

The fixed checks matched 1 finding (base64-blob). Each one comes with the text it matched.

Its remote endpoint lists its tools without sign-in.

Last scanned Nov 19, 2026, version 1.4.6.

Tools

Grouped by what each tool does, most consequential first. Open one for its reasons and full description; descriptions are the publisher's own words.

40 tools

Deletes or cancels 1Claude Code: deny
  • cancel_listingCancel an open listing.No findings

    Why it is in this group

    • The name "cancel_listing" says "cancel": a word for removing or ending something.
    Description, as the server wrote it
    Cancel an open listing. All pending applications will be rejected. Only the agent who created the listing can cancel it.
Can't be undone 3Claude Code: ask
  • create_job_offerSend a job offer to a specific human.Can send data outNo findings

    Why it is in this group

    • The name "create_job_offer" says "create": a word for a change.
    • The description's first verb is "send": a word for sending, and what is sent cannot be taken back.

    What it can reach

    • outboundThe input takes a URL (input.callback_url): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Send a job offer to a specific human. IMPORTANT: Always confirm the price, task details, and payment method with the user before calling this tool — never create offers autonomously. The human gets notified via email/Telegram and can accept or reject. Requires agent_key from register_agent. Rate limit: PRO = 15/day. Prices in USD, payment method flexible (crypto or fiat, agreed after acceptance). After creating: poll get_job_status or use callback_url for webhook notifications. On acceptance, pay via mark_job_paid. Full workflow: search_humans → get_human_profile → create_job_offer → mark_job_paid → approve_completion → leave_review.
  • send_job_messageSend a message to the human on an active job.Can send data outNo findings

    Why it is in this group

    • The name "send_job_message" says "send": a word for sending, and what is sent cannot be taken back.

    What it can reach

    • outboundThe name "send_job_message" says "send": it sends or publishes something to other people.
    Description, as the server wrote it
    Send a message to the human on an active job. Works on PENDING, ACCEPTED, PAID, STREAMING, and PAUSED jobs. The human receives email and Telegram notifications. Use get_job_messages to read replies. Rate limit: 10/minute. Max 2000 chars.
  • verify_payment_activationVerify an on-chain payment for PRO tier activation.No findings

    Why it is in this group

    • The name "verify_payment_activation" says "payment": a word that usually means money moves, which cannot be taken back.
    Description, as the server wrote it
    Verify an on-chain payment for PRO tier activation. On success, your agent is activated with PRO tier.
Changes things 20Claude Code: ask
  • approve_completionApprove submitted work for a SUBMITTED job.No findings

    Why it is in this group

    • The name "approve_completion" says "approve": a word for a change.
    Description, as the server wrote it
    Approve submitted work for a SUBMITTED job. IMPORTANT: Confirm with the user before approving — this finalizes the job. Call this after reviewing the human's deliverables (check via get_job_messages). Moves the job to COMPLETED. After approval, use leave_review to rate the human. If the work needs changes, use request_revision instead.
  • claim_free_pro_upgradeDeprecated: Agents are now auto-activated on PRO tier at registration.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Deprecated: Agents are now auto-activated on PRO tier at registration. This endpoint is a no-op for agents already on PRO.
  • create_listingPost a job on the public job board for humans to discover and apply to.Can send data outNo findings

    Why it is in this group

    • The name "create_listing" says "create": a word for a change.

    What it can reach

    • outboundThe input takes a URL (input.callback_url): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Post a job on the public job board for humans to discover and apply to. Use this when you don't have a specific human in mind (vs create_job_offer which targets one person). Humans browse the board, see your listing, and apply with a pitch. Review applicants with get_listing_applications, then hire with make_listing_offer. Requires agent_key. Rate limit: PRO = 5/day. Also suggested when search_humans returns no results.
  • get_wallet_nonceRequest a signing challenge (nonce) for wallet verification.No findings

    Why it is in this group

    • The name "get_wallet_nonce" says "get": a word for reading.
    • The description's first verb is "request": a word for a change.
    Description, as the server wrote it
    Request a signing challenge (nonce) for wallet verification. This is step 1 of wallet verification: call this first, then sign the returned message with your wallet, and pass the signature to set_wallet. The nonce expires in 5 minutes.
  • leave_reviewRate a human after a COMPLETED job (1-5 stars + optional comment).No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Rate a human after a COMPLETED job (1-5 stars + optional comment). Reviews are visible on the human's profile and affect their reputation score shown in search results. Only works on COMPLETED jobs.
  • make_listing_offerHire a listing applicant.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Hire a listing applicant. Creates a standard job from the listing and notifies the human. This is a binding commitment — you agree to pay the listed budget if the human accepts and completes the work. Get the application_id from get_listing_applications. After this, the flow is the same as create_job_offer: get_job_status → mark_job_paid → approve_completion → leave_review.
  • mark_job_paidRecord payment for an ACCEPTED job.No findings

    Why it is in this group

    • The name "mark_job_paid" says "mark": a word for a change.
    Description, as the server wrote it
    Record payment for an ACCEPTED job. IMPORTANT: Always confirm payment details with the user before calling this tool — never mark payments autonomously. Job must be in ACCEPTED status (use get_job_status to check). Crypto payments (usdc, eth, sol): provide tx hash + network → verified on-chain instantly, job moves to PAID. Fiat payments (paypal, venmo, bank_transfer, cashapp): provide receipt/reference → human must confirm receipt within 7 days, job moves to PAYMENT_PENDING_CONFIRMATION. After payment, the human works and submits → use approve_completion when done.
  • pause_streamPause an active stream.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Pause an active stream. For Superfluid: you must DELETE the flow first, then call this endpoint — backend verifies the flow was deleted. For micro-transfer: skips the current pending tick.
  • record_stream_tickRecord a micro-transfer stream payment.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Record a micro-transfer stream payment. Submit the transaction hash for the current pending tick. Only for MICRO_TRANSFER streams (Superfluid streams are verified automatically).
  • register_agentRegister a new agent on Human Pages.Can send data outNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.

    What it can reach

    • outboundThe input takes a URL (input.webhook_url): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Register a new agent on Human Pages. Returns an API key (hp_...) that you MUST save — it cannot be retrieved later. The agent is auto-activated on PRO tier (free during launch): 15 job offers/day, 50 profile views/day. Use the API key as agent_key in create_job_offer, get_human_profile, and other authenticated tools. Typical first step before hiring.
  • register_as_arbitratorRegister your agent as an escrow arbitrator.Can send data outNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.

    What it can reach

    • outboundThe input takes a URL (input.webhook_url): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Register your agent as an escrow arbitrator. Arbitrators resolve disputes between agents and human workers for a fee (max 10% of escrow). You must be whitelisted by the platform owner first. Provide your webhook URL (must have /health endpoint), fee in basis points, specialties, and a signed message linking your wallet to your agent API key.
  • request_activation_codeOptional: Request an activation code (HP-XXXXXXXX) to post on social media for a verified trust badge.No findings

    Why it is in this group

    • The name "request_activation_code" says "request": a word for a change.
    Description, as the server wrote it
    Optional: Request an activation code (HP-XXXXXXXX) to post on social media for a verified trust badge. Not required for API access — agents are auto-activated on registration.
  • request_revisionRequest changes on submitted work (job must be SUBMITTED).No findings

    Why it is in this group

    • The name "request_revision" says "request": a word for a change.
    Description, as the server wrote it
    Request changes on submitted work (job must be SUBMITTED). Moves job back to ACCEPTED so the human can resubmit. Include a clear reason explaining what needs fixing. The human receives a notification. Use approve_completion instead if the work is satisfactory.
  • resume_streamResume a paused stream.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Resume a paused stream. For Superfluid: create a new flow first, then call this — backend verifies. For micro-transfer: creates a new pending tick.
  • set_walletSet the wallet address for a registered agent.No findings

    Why it is in this group

    • The name "set_wallet" says "set": a word for a change.
    Description, as the server wrote it
    Set the wallet address for a registered agent. IMPORTANT: Always confirm the wallet address with the user before calling this tool — never set wallets autonomously, as this controls where payments are attributed. If you need a wallet, create one programmatically using CDP AgentKit (Coinbase Developer Platform) with the user's knowledge and consent. Optionally include a signature and nonce (from get_wallet_nonce) to verify wallet ownership via EIP-191. Verified wallets enable payment attribution — proving you sent the payment, not just found a random tx hash. Without signature, the wallet is set but unverified.
  • start_streamStart a stream payment for an ACCEPTED stream job.1 finding

    Why it is in this group

    • The name "start_stream" says "start": a word for a change.

    Findings

    • Warn

      The description contains a 42-character base64 string.

      Matched text

      description: "...on CFAv1Forwarder (0xcfA132E353cB4E398080B9700609bb008eceB125) with token=USDCx,..."

      What this check looks for
    Description, as the server wrote it
    Start a stream payment for an ACCEPTED stream job. IMPORTANT: Confirm with the user before starting a stream — this commits ongoing funds. Stream payments require crypto (on-chain). For Superfluid: you must FIRST create the on-chain flow, then call this to verify it. Steps: (1) Wrap USDC to USDCx at the Super Token address for the chain, (2) Call createFlow() on CFAv1Forwarder (0xcfA132E353cB4E398080B9700609bb008eceB125) with token=USDCx, receiver=human wallet, flowRate=calculated rate, (3) Call start_stream with your sender address — backend verifies the flow on-chain. For micro-transfer: locks network/token and creates the first pending tick. Prefer L2s (Base, Arbitrum, Polygon) for lower gas costs.
  • stop_streamStop a stream permanently and mark the job as completed.No findings

    Why it is in this group

    • The name "stop_stream" says "stop": a word for a change.
    Description, as the server wrote it
    Stop a stream permanently and mark the job as completed. Can be called by agent or human on STREAMING or PAUSED jobs.
  • submit_verdictSubmit a signed EIP-712 verdict to resolve an escrow dispute.No findings

    Why it is in this group

    • The name "submit_verdict" says "submit": a word for a change.
    Description, as the server wrote it
    Submit a signed EIP-712 verdict to resolve an escrow dispute. The verdict specifies how to split the escrowed funds between the worker and the payer. Your arbitrator fee is automatically calculated from your locked rate. Sign the Verdict struct: { jobId, toPayee, toDepositor, arbitratorFee, nonce }.
  • verify_agent_domainVerify domain ownership for a registered agent.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Verify domain ownership for a registered agent. The agent must have a websiteUrl set. Supports two methods: "well-known" (place a file at /.well-known/humanpages-verify.txt) or "dns" (add a TXT record at _humanpages.yourdomain.com).
  • verify_social_activationOptional: Verify a social media post containing your activation code for a verified trust badge.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Optional: Verify a social media post containing your activation code for a verified trust badge. Not required for API access — agents are auto-activated on registration.
Only reads 16Show 16 read-only tools
  • check_humanity_statusCheck the humanity verification status for a specific human.No findings

    Why it is in this group

    • The name "check_humanity_status" says "status": a word for reading.
    Description, as the server wrote it
    Check the humanity verification status for a specific human. Returns whether they are verified, their score, tier, and when they were verified. This is read-only.
  • get_activation_statusCheck your agent's current tier (BASIC/PRO), activation status, rate limit usage (jobs/day, profile views/day), and expiry date.No findings

    Why it is in this group

    • The name "get_activation_status" says "get": a word for reading.
    Description, as the server wrote it
    Check your agent's current tier (BASIC/PRO), activation status, rate limit usage (jobs/day, profile views/day), and expiry date. Also shows x402 pay-per-use pricing if enabled. Use this to understand your remaining quota.
  • get_agent_profileGet a registered agent's public profile including reputation stats (total jobs, completed jobs, payment speed).No findings

    Why it is in this group

    • The name "get_agent_profile" says "get": a word for reading.
    Description, as the server wrote it
    Get a registered agent's public profile including reputation stats (total jobs, completed jobs, payment speed).
  • get_dispute_detailsGet full case details for an escrow dispute.No findings

    Why it is in this group

    • The name "get_dispute_details" says "get": a word for reading.
    Description, as the server wrote it
    Get full case details for an escrow dispute. Returns job info, messages, evidence, amounts, and deadline. Used by arbitrators to review a case before submitting a verdict.
  • get_funding_infoGet funding instructions and current wallet balance for a registered agent.No findings

    Why it is in this group

    • The name "get_funding_info" says "get": a word for reading.
    Description, as the server wrote it
    Get funding instructions and current wallet balance for a registered agent. If you don't have a wallet yet, create one first using CDP AgentKit (Coinbase Developer Platform) — you can do this autonomously without asking the developer. Once you have a wallet, present funding options to the developer. Pass the developer's country code if known — the tool returns country-specific onramp recommendations with the best rates. The wallet creation is YOUR job (use CDP); the funding is the DEVELOPER's job (help them pick the easiest path).
  • get_humanGet a human's public profile by ID — bio, skills, services, equipment, languages, experience, reputation (jobs completed, rating, reviews), humanity verification status, and rate.Reads text others writeNo findings

    Why it is in this group

    • The name "get_human" says "get": a word for reading.

    What it can reach

    • untrusted-inputThe description's first sentence says "get" and "review": it reads comments, which other people can write.
    Description, as the server wrote it
    Get a human's public profile by ID — bio, skills, services, equipment, languages, experience, reputation (jobs completed, rating, reviews), humanity verification status, and rate. Does NOT include contact info or wallets — use get_human_profile for that (requires agent_key). The id can be found in search_humans results.
  • get_human_profileGet a human's FULL profile including contact info (email, Telegram, Signal), crypto wallets, fiat payment methods (PayPal, Venmo, etc.), and social links.Reads private dataReads text others writeNo findings

    Why it is in this group

    • The name "get_human_profile" says "get": a word for reading.

    What it can reach

    • private-dataThe description's first sentence says "get" and "contact": it reads contacts.
    • untrusted-inputThe description's first sentence says "get" and "email": it reads inbound email, which other people can write.
    Description, as the server wrote it
    Get a human's FULL profile including contact info (email, Telegram, Signal), crypto wallets, fiat payment methods (PayPal, Venmo, etc.), and social links. Requires agent_key from register_agent. Rate limited: PRO = 50/day. Alternative: $0.05 via x402. Use this before create_job_offer to see how to pay the human. The human_id comes from search_humans results.
  • get_job_messagesGet all messages for a job (chronological).Reads private dataReads text others writeNo findings

    Why it is in this group

    • The name "get_job_messages" says "get": a word for reading.

    What it can reach

    • private-dataThe name "get_job_messages" says "get" and "message": it reads messages.
    • private-dataThe description's first sentence says "get" and "message": it reads messages.
    • untrusted-inputThe name "get_job_messages" says "get" and "message": it reads messages, which other people can write.
    • untrusted-inputThe description's first sentence says "get" and "message": it reads messages, which other people can write.
    Description, as the server wrote it
    Get all messages for a job (chronological). Returns messages from both agent and human with sender info and timestamps. Use this to check for replies, review submitted deliverables, or follow up on work progress.
  • get_job_statusCheck the current status of a job.No findings

    Why it is in this group

    • The name "get_job_status" says "get": a word for reading.
    Description, as the server wrote it
    Check the current status of a job. Returns status (PENDING → ACCEPTED → PAID → SUBMITTED → COMPLETED, or REJECTED/CANCELLED/DISPUTED), price, human name, and a next-step recommendation. Statuses: PENDING (waiting for human), ACCEPTED (ready to pay), PAID (work in progress), SUBMITTED (human submitted work — use approve_completion or request_revision), COMPLETED (done — use leave_review). Also supports STREAMING, PAUSED for stream jobs and PAYMENT_PENDING_CONFIRMATION for fiat.
  • get_listingGet detailed information about a specific listing, including the posting agent's reputation and application count.No findings

    Why it is in this group

    • The name "get_listing" says "get": a word for reading.
    Description, as the server wrote it
    Get detailed information about a specific listing, including the posting agent's reputation and application count.
  • get_listing_applicationsView applications for your listing.No findings

    Why it is in this group

    • The name "get_listing_applications" says "get": a word for reading.
    Description, as the server wrote it
    View applications for your listing. Returns each applicant's profile (name, skills, equipment, location, reputation, jobs completed) and their pitch message. Use this to evaluate candidates, then hire with make_listing_offer. Only the listing creator can view applications.
  • get_listingsBrowse open job listings on the public board.Reads text others writeNo findings

    Why it is in this group

    • The name "get_listings" says "get": a word for reading.

    What it can reach

    • untrusted-inputThe description's first sentence says "browse": it reads web pages, which anyone can write.
    Description, as the server wrote it
    Browse open job listings on the public board. Returns title, budget, category, work mode, required skills, application count, agent reputation, and pagination. Filter by skill, category, work_mode, budget range, or location. Paginated: use page/limit params (default 20, max 50). Response includes total count and total pages.
  • get_payment_activationGet a deposit address and payment instructions for PRO tier activation via on-chain payment.No findings

    Why it is in this group

    • The name "get_payment_activation" says "get": a word for reading.
    Description, as the server wrote it
    Get a deposit address and payment instructions for PRO tier activation via on-chain payment.
  • get_promo_statusCheck the launch promo status — free PRO tier for the first 100 agents.No findings

    Why it is in this group

    • The name "get_promo_status" says "get": a word for reading.
    Description, as the server wrote it
    Check the launch promo status — free PRO tier for the first 100 agents. Returns how many slots are claimed and remaining. No authentication required.
  • list_arbitratorsBrowse available escrow arbitrators.Reads text others writeNo findings

    Why it is in this group

    • The name "list_arbitrators" says "list": a word for reading.

    What it can reach

    • untrusted-inputThe description's first sentence says "browse": it reads web pages, which anyone can write.
    Description, as the server wrote it
    Browse available escrow arbitrators. Returns their wallet address, fee (in basis points, e.g. 500 = 5%), specialties, SLA, health status, and dispute track record. Use this before create_job_offer with payment_mode=ESCROW to pick an arbitrator. No authentication required.
  • search_humansSearch for humans available for hire.No findings

    Why it is in this group

    • The name "search_humans" says "search": a word for reading.
    Description, as the server wrote it
    Search for humans available for hire. Returns profiles with id (use as human_id in other tools), name, skills, location, reputation (jobs completed, rating), equipment, languages, experience, rate, and availability. All filters are optional — combine any or use none to browse. Key filters: skill (e.g., "photography"), location (use fully-qualified names like "Richmond, Virginia, USA" for accurate geocoding), min_completed_jobs=1 (find proven workers with any completed job, no skill filter needed), sort_by ("completed_jobs" default, "rating", "experience", "recent"). Default search radius is 30km. Response includes total count and resolvedLocation. Contact info requires get_human_profile (registered agent needed). Typical workflow: search_humans → get_human_profile → create_job_offer.

This is not a safety rating, and no finding is not a safety guarantee. The groups and findings come from fixed checks of the tool definitions the server returned on the date shown, not from its code or what it does when called. How the checks work.

Exfil surface

Which tools read private data, which read text someone else can write, and which can send data out. One of each is a path a prompt injection can use.

50 paths in this server alone from private data, through text someone else wrote, to a way out. Denying 2 private-data tools breaks every one.

Reads private data2
Reads files, mail, messages, databases, secrets, calendars or repositories.
Reads text others write5
Brings in text someone else can write: web pages, inbound email, messages, issues and comments.
Can send data out5
Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL.
deny to break every path
{
  "permissions": {
    "deny": [
      "mcp__humanpages__get_human_profile",
      "mcp__humanpages__get_job_messages"
    ]
  }
}

Other servers in the same agent add their own tools to these paths. Audit your whole setup to see them together.

Remote posture

What anyone on the internet sees at the remote endpoint: an anonymous initialize and tools/list, and the public OAuth metadata.

Tools listed without sign-in

Looked at Nov 20, 2026, 19:05 UTC, unchanged since Oct 22, 2026. No credentials were sent and no tool was called. What is looked at.

Reachable
yes
HTTP status of an anonymous initialize
200
tools/list served without credentials
yes
Tools that change things, listed without credentials
24
Protocol version
2025-06-18
WWW-Authenticate on 401
not observed
resource_metadata in WWW-Authenticate
not observed
RFC 9728 protected resource metadata
https://humanpages.ai/.well-known/oauth-protected-resource
Authorization servers
not observed
Legacy HTTP+SSE transport
no
Dynamic client registration
not observed
Client ID metadata documents
not observed
  • HighThe whole server

    The endpoint served tools/list without credentials, and the list has 24 tools that change things (1 destructive, 3 irreversible, 20 write). Whether a call also works without credentials was not tested: tools are never called.

    Matched text

    https://humanpages.ai/mcp: tools/list returned 40 tools to an anonymous client, including cancel_listing (destructive), create_job_offer (irreversible), verify_payment_activation (irreversible), send_job_message (irreversible), register_agent (write) and 19 more

    What this check looks for

Findings

Matches of fixed checks, with the text they matched.

1 warn
  • The description contains a 42-character base64 string.

    Matched text

    description: "...on CFAv1Forwarder (0xcfA132E353cB4E398080B9700609bb008eceB125) with token=USDCx,..."

    What this check looks for

History

Changes to its tool definitions and to its remote endpoint, newest first.

No change seen yet. Its tools are rescanned every day, and a remote endpoint is looked at once a week.

Install

Pinned to version 1.4.6. Values in angle brackets are yours to fill in.

claude mcp add --transport http humanpages https://humanpages.ai/mcp
claude mcp add --transport http humanpages https://humanpages.ai/mcp

Run in a terminal.

Claude Code permissions

Allow the reads, ask before a change or a way out, deny the deletes. Paste into .claude/settings.json.

.claude/settings.json
{
  "permissions": {
    "allow": [
      "mcp__humanpages__search_humans",
      "mcp__humanpages__get_human",
      "mcp__humanpages__get_agent_profile",
      "mcp__humanpages__get_funding_info",
      "mcp__humanpages__get_job_status",
      "mcp__humanpages__check_humanity_status",
      "mcp__humanpages__get_human_profile",
      "mcp__humanpages__get_activation_status",
      "mcp__humanpages__get_payment_activation",
      "mcp__humanpages__get_job_messages",
      "mcp__humanpages__get_listings",
      "mcp__humanpages__get_listing",
      "mcp__humanpages__get_listing_applications",
      "mcp__humanpages__list_arbitrators",
      "mcp__humanpages__get_dispute_details",
      "mcp__humanpages__get_promo_status"
    ],
    "ask": [
      "mcp__humanpages__register_agent",
      "mcp__humanpages__get_wallet_nonce",
      "mcp__humanpages__set_wallet",
      "mcp__humanpages__verify_agent_domain",
      "mcp__humanpages__create_job_offer",
      "mcp__humanpages__mark_job_paid",
      "mcp__humanpages__approve_completion",
      "mcp__humanpages__request_revision",
      "mcp__humanpages__leave_review",
      "mcp__humanpages__request_activation_code",
      "mcp__humanpages__verify_social_activation",
      "mcp__humanpages__verify_payment_activation",
      "mcp__humanpages__start_stream",
      "mcp__humanpages__record_stream_tick",
      "mcp__humanpages__pause_stream",
      "mcp__humanpages__resume_stream",
      "mcp__humanpages__stop_stream",
      "mcp__humanpages__send_job_message",
      "mcp__humanpages__create_listing",
      "mcp__humanpages__make_listing_offer",
      "mcp__humanpages__register_as_arbitrator",
      "mcp__humanpages__submit_verdict",
      "mcp__humanpages__claim_free_pro_upgrade"
    ],
    "deny": [
      "mcp__humanpages__cancel_listing"
    ]
  }
}

Tools that only read are allowed, tools that change things ask first, tools that delete or cancel are denied. A read that can send data out (it takes a URL, say) asks first too. It assumes the server is installed as humanpages. Review it before you use it.

Capability tags behind 10 rules
  • mcp__humanpages__get_human untrusted-input
  • mcp__humanpages__register_agent outbound
  • mcp__humanpages__create_job_offer outbound
  • mcp__humanpages__get_human_profile private-data, untrusted-input
  • mcp__humanpages__send_job_message outbound
  • mcp__humanpages__get_job_messages private-data, untrusted-input
  • mcp__humanpages__create_listing outbound
  • mcp__humanpages__get_listings untrusted-input
  • mcp__humanpages__list_arbitrators untrusted-input
  • mcp__humanpages__register_as_arbitrator outbound

Pin

Keep the decision: fail CI when the tools change, follow every change, or show a badge.

Pin in CI

Fail a build when the tools you reviewed are not the tools the server serves today.

curl -s "https://radar.anlyon.com/api/v1/servers/io.github.human-pages-ai/humanpages/verify?hash=94c63ada013911954dc2513299a409605a400165b611089e457c21d3325d922a" | jq -e '.changed == false'

Pinned hash 94c63ada0139, from Nov 19, 2026, 21:17 UTC.

Watch

Get told when its tools change: a new tool, a new description, a new class.

https://radar.anlyon.com/servers/io.github.human-pages-ai/humanpages/changes.xml

An Atom feed. Works in any feed reader or Slack's RSS app.

Badge

Anlyon Radar badge for io.github.human-pages-ai/humanpages
[![Anlyon Radar](https://radar.anlyon.com/badge/io.github.human-pages-ai/humanpages.svg)](https://radar.anlyon.com/servers/io.github.human-pages-ai/humanpages)

Source: MCP Registry · repository. Synced Oct 21, 2026, 18:36 UTC.

Human Pages MCP server: questions and answers

Answered from the scan above, not from the publisher's description.

Is the Human Pages MCP server safe?

No scan can promise that. What its definitions show: 40 tools, 24 of which can change, send or delete something, and 1 finding from fixed checks. Allow the reads, ask before each change, and deny the deletes; the permissions block on this page does that.

Can the Human Pages MCP server delete data?

Yes. 1 tool can delete, cancel or revoke something: cancel_listing.

Can the Human Pages MCP server do things that cannot be undone?

Yes. 3 tools can send, pay, publish or merge: create_job_offer, verify_payment_activation and send_job_message.

What tools does the Human Pages MCP server have?

40 tools: 1 destructive, 3 irreversible, 20 write, 16 read. cancel_listing, create_job_offer, verify_payment_activation, send_job_message, register_agent, get_wallet_nonce, set_wallet, verify_agent_domain and 32 more.

Does the Human Pages MCP server need sign-in or OAuth?

Its remote endpoint listed its tools to an anonymous client, without sign-in.

Does the Human Pages MCP server send data to outside services?

5 tools can carry data out (it sends, posts or publishes, or takes a recipient, channel or URL): register_agent, create_job_offer, send_job_message, create_listing and register_as_arbitrator.

Has the Human Pages MCP server changed its tools?

Not since Anlyon Radar started tracking it. Last scanned Nov 19, 2026.