Skip to content

HumanEndpoint MCP server

au.humanendpoint/humanendpointhumanendpoint.austreamable-httpv1.0.1Scanned

Human-only capabilities for AI agents: ground truth, verification, judgement and real-world action.

10 tools: 2 can delete or cancel things, 1 do something that can't be undone, 2 change things and 5 only read.

Tools: 2 delete, 1 irreversible, 2 write, 5 read.
  • Reads private data · 1
  • Reads text others write · 5
  • Can send data out · 2
  • No findings

Tools listed without sign-in. Remote posture

With Anlyon

2 tools here can delete or cancel things and 1 does something that can't be undone. Govern them.

  • Anlyon lists the tools and adopts none.
  • Each tool you adopt gets approvals and receipts.
  • Your agent never holds the key.
Govern this server →

A tool your agent calls directly is not routed through Anlyon and is not governed by it. Anlyon Radar only lists and labels, and these labels are the same either way.

What the HumanEndpoint MCP server can do

The HumanEndpoint MCP server is published by humanendpoint.au, a domain the MCP Registry verified. It has 10 tools. It can delete or cancel things (2 tools: human_request_quote and human_cancel_job), do things that cannot be undone, such as sending, paying or publishing (human_reply_to_job) and create or change data (2 tools: human_policy_preflight and human_create_job). 5 tools only read.

1 tool reads private data (human_get_job); 5 tools bring in text other people can write (human_list_services, human_create_job, human_get_job and 2 more); 2 tools can send data out of your system (human_create_job and human_reply_to_job). On its own it has all three parts of a prompt-injection exfiltration path: private data, text others write, and a way out.

None of the fixed checks matched anything in its tool definitions. That is not a safety guarantee.

Its remote endpoint lists its tools without sign-in, and its tools changed once since Anlyon Radar first saw them.

Last scanned Nov 12, 2026, version 1.0.1.

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.

10 tools

Deletes or cancels 2Claude Code: deny
  • human_cancel_jobCancel an unpaid HumanEndpoint job immediately, or request cancellation of human work that may already be underway and may require refund or operator handling.Reads text others writeNo findings

    Why it is in this group

    • The name "human_cancel_job" says "cancel": a word for removing or ending something.
    • The server marks it destructive (destructiveHint: true).

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cancel an unpaid HumanEndpoint job immediately, or request cancellation of human work that may already be underway and may require refund or operator handling.

    Annotations the server declares

    • readOnlyHint: false
    • openWorldHint: true
    • destructiveHint: true
  • human_request_quoteSubmit a request for human-reviewed terms for a real-world task whose cost depends on scope, feasibility, urgency or conditions.No findings

    Why it is in this group

    • The name "human_request_quote" says "request": a word for a change.
    • The server marks it destructive (destructiveHint: true).
    Description, as the server wrote it
    Submit a request for human-reviewed terms for a real-world task whose cost depends on scope, feasibility, urgency or conditions. This sends the request into a HumanEndpoint workflow for review by a human operator, but does not commit funds or start execution. HumanEndpoint returns commercial terms that can be accepted or declined before payment. Provide callback_url to receive signed quote.offered or quote.declined webhooks instead of relying only on polling. Suitable tasks include calling a business for current price, stock, lead time or availability; retrieving information only available by phone; verifying a real-world fact; checking AI output; applying human judgement; doing bounded research; or carrying out a physical-world task. If no named service matches, set service_slug to custom-human-task.

    Annotations the server declares

    • readOnlyHint: false
    • openWorldHint: false
    • destructiveHint: true
Can't be undone 1Claude Code: ask
  • human_reply_to_jobSend additional information to the human operator working on the job, including when the job is waiting for agent input.Reads text others writeCan send data outNo findings

    Why it is in this group

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

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • outboundThe name "human_reply_to_job" says "reply": it sends or publishes something to other people.
    Description, as the server wrote it
    Send additional information to the human operator working on the job, including when the job is waiting for agent input.

    Annotations the server declares

    • readOnlyHint: false
    • openWorldHint: true
    • destructiveHint: false
Changes things 2Claude Code: ask
  • human_create_jobAccept the human-reviewed commercial terms using the quote_token and create the asynchronous HumanEndpoint job.Reads text others writeCan send data outNo findings

    Why it is in this group

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

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • 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
    Accept the human-reviewed commercial terms using the quote_token and create the asynchronous HumanEndpoint job. Creating the job starts a workflow that may lead to human execution in the real world after required payment. Store the returned job_token because it is required for later reads and control.

    Annotations the server declares

    • readOnlyHint: false
    • openWorldHint: true
    • destructiveHint: false
  • human_policy_preflightCheck whether a proposed human task is eligible before requesting a quote or spending funds.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.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Check whether a proposed human task is eligible before requesting a quote or spending funds.

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: false
    • destructiveHint: false
Only reads 5Show 5 read-only tools
  • human_get_jobRetrieve job status, messages, history, payment state, expense accounting, result and evidence.Reads private dataReads text others writeNo findings

    Why it is in this group

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

    What it can reach

    • private-dataThe description's first sentence says "retrieve" and "message": it reads messages.
    • untrusted-inputThe description's first sentence says "retrieve" and "message": it reads messages, which other people can write.
    Description, as the server wrote it
    Retrieve job status, messages, history, payment state, expense accounting, result and evidence.

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: false
    • destructiveHint: false
  • human_get_quoteRetrieve a human-reviewed quote.No findings

    Why it is in this group

    • The name "human_get_quote" says "get": a word for reading.
    Description, as the server wrote it
    Retrieve a human-reviewed quote. Polling is always available as a fallback even when the quote request supplied callback_url. When status=offered, inspect the operator's service fee, expense authority and expiry. Creating the job accepts those terms. When declined, inspect the reason before deciding whether to resubmit.

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: false
    • destructiveHint: false
  • human_list_evidenceList evidence artifacts and SHA-256 hashes for a secured human job.No findings

    Why it is in this group

    • The name "human_list_evidence" says "list": a word for reading.
    Description, as the server wrote it
    List evidence artifacts and SHA-256 hashes for a secured human job. Download paths require the same job token.

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: false
    • destructiveHint: false
  • human_list_expensesList pass-through expense requests for a secured job, including exact amount, payment status and receipt metadata.No findings

    Why it is in this group

    • The name "human_list_expenses" says "list": a word for reading.
    Description, as the server wrote it
    List pass-through expense requests for a secured job, including exact amount, payment status and receipt metadata.

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: false
    • destructiveHint: false
  • human_list_servicesDiscover human capabilities for tasks software cannot complete alone: call a business, supplier or organisation; check current price, stock, lead time or availability by phone; obt…Reads text others writeNo findings

    Why it is in this group

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

    What it can reach

    • untrusted-inputThe description's first sentence says "review": it reads comments, which other people can write.
    Description, as the server wrote it
    Discover human capabilities for tasks software cannot complete alone: call a business, supplier or organisation; check current price, stock, lead time or availability by phone; obtain information that is only available by speaking to a person; verify a physical fact or location in the real world; independently review AI output; apply human judgement; perform bounded manual research; or carry out another lawful physical-world action. Results are structured and can include evidence. If no listed capability fits, use custom-human-task.

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: false
    • destructiveHint: false

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.

10 paths in this server alone from private data, through text someone else wrote, to a way out. Denying 1 private-data tool breaks every one.

Reads private data1
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 out2
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__humanendpoint__human_get_job"
    ]
  }
}

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, 18:31 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
5
Protocol version
2025-06-18
WWW-Authenticate on 401
not observed
resource_metadata in WWW-Authenticate
not observed
RFC 9728 protected resource metadata
not observed
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 5 tools that change things (2 destructive, 1 irreversible, 2 write). Whether a call also works without credentials was not tested: tools are never called.

    Matched text

    https://humanendpoint.au/mcp: tools/list returned 10 tools to an anonymous client, including human_request_quote (destructive), human_cancel_job (destructive), human_reply_to_job (irreversible), human_policy_preflight (write), human_create_job (write)

    What this check looks for

Findings

Matches of fixed checks, with the text they matched.

No check matched anything in these tool definitions. That is not a safety guarantee.

History

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

  1. Tools2 changed.~2 changedView diff

Install

Pinned to version 1.0.1. Values in angle brackets are yours to fill in.

claude mcp add --transport http humanendpoint https://humanendpoint.au/mcp
claude mcp add --transport http humanendpoint https://humanendpoint.au/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__humanendpoint__human_list_services",
      "mcp__humanendpoint__human_get_quote",
      "mcp__humanendpoint__human_get_job",
      "mcp__humanendpoint__human_list_expenses",
      "mcp__humanendpoint__human_list_evidence"
    ],
    "ask": [
      "mcp__humanendpoint__human_policy_preflight",
      "mcp__humanendpoint__human_create_job",
      "mcp__humanendpoint__human_reply_to_job"
    ],
    "deny": [
      "mcp__humanendpoint__human_request_quote",
      "mcp__humanendpoint__human_cancel_job"
    ]
  }
}

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 humanendpoint. Review it before you use it.

Capability tags behind 5 rules
  • mcp__humanendpoint__human_list_services untrusted-input
  • mcp__humanendpoint__human_create_job untrusted-input, outbound
  • mcp__humanendpoint__human_get_job private-data, untrusted-input
  • mcp__humanendpoint__human_reply_to_job untrusted-input, outbound
  • mcp__humanendpoint__human_cancel_job untrusted-input

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/au.humanendpoint/humanendpoint/verify?hash=fa82bab272c139738f694feabcd00201527f3091e0670d898a2499cd0f2e1294" | jq -e '.changed == false'

Pinned hash fa82bab272c1, from Nov 12, 2026, 05:18 UTC.

Watch

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

https://radar.anlyon.com/servers/au.humanendpoint/humanendpoint/changes.xml

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

Badge

Anlyon Radar badge for au.humanendpoint/humanendpoint
[![Anlyon Radar](https://radar.anlyon.com/badge/au.humanendpoint/humanendpoint.svg)](https://radar.anlyon.com/servers/au.humanendpoint/humanendpoint)

Source: MCP Registry. Synced Oct 21, 2026, 17:45 UTC.

HumanEndpoint MCP server: questions and answers

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

Is the HumanEndpoint MCP server safe?

No scan can promise that. What its definitions show: 10 tools, 5 of which can change, send or delete something, and 0 findings from fixed checks. Allow the reads, ask before each change, and deny the deletes; the permissions block on this page does that.

Can the HumanEndpoint MCP server delete data?

Yes. 2 tools can delete, cancel or revoke something: human_request_quote and human_cancel_job.

Can the HumanEndpoint MCP server do things that cannot be undone?

Yes. 1 tool can send, pay, publish or merge: human_reply_to_job.

What tools does the HumanEndpoint MCP server have?

10 tools: 2 destructive, 1 irreversible, 2 write, 5 read. human_request_quote, human_cancel_job, human_reply_to_job, human_policy_preflight, human_create_job, human_list_services, human_get_quote, human_get_job and 2 more.

Does the HumanEndpoint MCP server need sign-in or OAuth?

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

Does the HumanEndpoint MCP server send data to outside services?

2 tools can carry data out (it sends, posts or publishes, or takes a recipient, channel or URL): human_create_job and human_reply_to_job.

Has the HumanEndpoint MCP server changed its tools?

Yes, once since Anlyon Radar started tracking it, most recently on Nov 12, 2026. Every change is on this page and in its Atom feed.