Skip to content

Agent Control — Action Gate MCP server

net.agent-control/agent-controlagent-control.netstreamable-httpv1.0.5Scanned

Agent asks a person go/stop/wait before Slack, CRM or risky steps. $49 for 30 days.

26 tools: 4 do something that can't be undone, 16 change things and 6 only read.

Tools: 0 delete, 4 irreversible, 16 write, 6 read.
  • Reads private data · 0
  • Reads text others write · 3
  • Can send data out · 7
  • 1 finding

Tools listed without sign-in. Remote posture

With Anlyon

4 tools here do 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 Agent Control — Action Gate MCP server can do

The Agent Control — Action Gate MCP server is published by agent-control.net, a domain the MCP Registry verified. It has 26 tools. It can do things that cannot be undone, such as sending, paying or publishing (meter_preflight, meter_stamp, check_transfer and start_trial) and create or change data (16 tools: post_job, list_your_agent, x402_listing_check, agent_readiness_check and 12 more), but cannot delete data. 6 tools only read.

3 tools bring in text other people can write (list_open_jobs, list_your_agent and browse_agents); 7 tools can send data out of your system (post_job, x402_listing_check, agent_readiness_check and 4 more).

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

Its remote endpoint lists its tools without sign-in, and its tools changed 2 times since Anlyon Radar first saw them.

Last scanned Nov 8, 2026, version 1.0.5.

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.

26 tools

Can't be undone 4Claude Code: ask
  • check_transferMUST be called before the agent signs or sends a transfer.Can send data outNo findings

    Why it is in this group

    • The name "check_transfer" says "transfer": a word for a change that another call does not undo.
    • The input addresses people (input.to): what it sends cannot be taken back.

    What it can reach

    • outboundThe input takes a destination (input.to): what the tool writes goes there.
    Description, as the server wrote it
    MUST be called before the agent signs or sends a transfer. Returns allow, alert, hold, or block. If must_abort is true, do not sign. If decision is hold, poll get_approval until allow or block.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
  • meter_preflightCan I pay this address?Can send data outNo findings

    Why it is in this group

    • The input addresses people (input.to): what it sends cannot be taken back.

    What it can reach

    • outboundThe input takes a destination (input.to): what the tool writes goes there.
    Description, as the server wrote it
    Can I pay this address? Buy looks_20 pack ($0.20) or look $0.10. Stamp ticket $0.05. No inbox. One look = one address. No free looks. Without a paid X-Agent-Pass, 402 looks_20 pack $0.20. look $0.10 is optional one-shot. Body: chain, wallet, to, value_usd, cap_usd. allow or stop vs cap_usd. Never hold.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
  • meter_stampCan I pay this address?Can send data outNo findings

    Why it is in this group

    • The input addresses people (input.to): what it sends cannot be taken back.

    What it can reach

    • outboundThe input takes a destination (input.to): what the tool writes goes there.
    Description, as the server wrote it
    Can I pay this address? Buy looks_20 pack ($0.20) or look $0.10. Stamp ticket $0.05. No inbox. stamp_tx $0.05. Take this ticket or we do not take your USDC. Signed allow|stop receipt. HMAC-SHA256. Merchants verify with meter_verify_stamp before accepting agent USDC.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
  • start_trialInvite a human to start the 1-day trial.No findings

    Why it is in this group

    • The name "start_trial" says "start": a word for a change.
    • The description's first verb is "invite": a word for sending, and what is sent cannot be taken back.
    Description, as the server wrote it
    Invite a human to start the 1-day trial. Provide human_email or an existing principal_id. Agents cannot open a root account. The human owns billing and Approval Inbox.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
Changes things 16Claude Code: ask
  • agent_readiness_checkFree.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.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Free. No API key. Read a public https site and return the first five checks for robots.txt, llms.txt, agents.json, descriptions, and skill.md. Same as POST /api/v1/agent-check. A pass does not mean a crawler or an agent will use the site.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
  • ask_humanSame as check_action (Ask).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
    Same as check_action (Ask). Stop and ask before a consequential non-money action. Returns go, stop, or wait plus approval_id. Poll get_approval. Timeout is stop. Does not forward Slack or CRM. Use slack.post or crm.write to enforce. Email isn't supported yet.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
  • attach_humanAttach this agent to a human principal (email or existing principal_id).No findings

    Why it is in this group

    • The name "attach_human" says "attach": a word for a change.
    Description, as the server wrote it
    Attach this agent to a human principal (email or existing principal_id). Does not move the agent to a different human. Agents cannot decide Approval Inbox.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
  • check_actionAction Gate Ask.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
    Action Gate Ask. MUST be called before a consequential non-money action when the agent still executes after go (including deploy). Returns go, stop, or wait. If decision is wait, poll get_approval until go or stop. No decision in 10 minutes = stop. If must_abort is true, do not act. This tool does not send the post or CRM write. To enforce those, call slack.post or crm.write. Email isn't supported yet. Alias: ask_human.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
  • create_checkoutOpen a pay request on the human principal that owns this agent.No findings

    Why it is in this group

    • The name "create_checkout" says "create": a word for a change.
    Description, as the server wrote it
    Open a pay request on the human principal that owns this agent. Wraps POST /api/v1/billing/checkout. plan action opens Action Gate at $49 USDC. starter, pro, and team open the wallet console. The human pays at pay_url in USDC on Base, Solana or Ethereum. Not automatic payment. Agents cannot decide Approval Inbox.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
  • crm.writeWrite Gate.No findings

    Why it is in this group

    • The name "crm.write" says "write": a word for a change.
    Description, as the server wrote it
    Write Gate. Validates the preview, runs Action Gate check_action for crm.write, and POSTs the approved preview to the human's CRM webhook only when the decision is go. Unpaid, missing DATABASE_URL, stop, and timeout do not post. If decision is wait, poll get_approval, then call crm.write again with approval_id. If the CRM webhook is missing, returns stop and does not post. Same $49 Action Gate seat (plan=action).

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
  • list_your_agentFree job board: https://agent-control.net/exchange.Reads text others write1 finding

    Why it is in this group

    • The name "list_your_agent" says "list": a word for reading.
    • The server marks it as not read-only (readOnlyHint: false).

    What it can reach

    • untrusted-inputThe description's first sentence says "http": it reads web pages, which anyone can write.

    Findings

    • Warn

      The server marks "list_your_agent" as not read-only (readOnlyHint: false), and its words say it reads.

      Matched text

      annotations: {"readOnlyHint":false,"destructiveHint":false}; The name "list_your_agent" says "list": a word for reading.

      What this check looks for
    Description, as the server wrote it
    Free job board: https://agent-control.net/exchange. Posting costs nothing. Featured is $19 for 7 days and pins a listing on top. Hire paths pay us directly. Free agent list: https://agent-control.net/directory. Listing your agent is free. No API key. No Meter look. No Action Gate seat. Add your agent to the free directory. Listing your agent is free. People reach you at the contact you leave. Same fields as POST /api/v1/agents/listings. Skill: https://agent-control.net/skill.md

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
  • meter_buy_passCan I pay this address?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
    Can I pay this address? Buy looks_20 pack ($0.20) or look $0.10. Stamp ticket $0.05. No inbox. Call with sku looks_20 (or omit) to buy a pack that mints/extends X-Agent-Pass. look $0.10 is optional one-shot. Returns a payable invoice as tool content (ok:true / status payment_required) with pay_to, base_pay_to, accepts (Base USDC EIP-3009 exact and Solana USDC), amount_usd, amount_base_units, reference, pay_url, invoice_id, watch_url, adapter_url, base_adapter_url, preferred_rail, pay_page, next_tool, tool_fields, sign_exact, adapter_snippet, next_steps. Prefer Base EIP-3009 exact to base_pay_to (CDP/AgentKit; no Solana key) then meter_watch with payment. Use sign_exact + adapter_snippet to finish in-process (payMeterPassBase / signTypedData). Fetch base_adapter_url only if you need the full file. Optional Solana: fetch adapter_url. 402 invoice includes pay_page for Phantom laptop as a last resort. We never take keys. Packs: looks_20 / addresses_100 / stamp_tx. No human account.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
  • meter_pricingCan I pay this address?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
    Can I pay this address? Buy looks_20 pack ($0.20) or look $0.10. Stamp ticket $0.05. No inbox. No Authorization / Bearer empty. Public catalog from this tool: look, looks_20, addresses_100, stamp_tx. default_sku is look. paid_sku looks_20. Base USDC (EIP-3009 exact) and Solana USDC. No email. No API key. Human App $29 plans are separate.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
  • meter_scanCan I pay this address?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
    Can I pay this address? Buy looks_20 pack ($0.20) or look $0.10. Stamp ticket $0.05. No inbox. One look = one address. No free looks. Without a paid X-Agent-Pass, 402 looks_20 pack $0.20 (look $0.10 is optional one-shot). Prefer Base USDC (EIP-3009 exact) to base_pay_to (CDP/AgentKit; no Solana key). Optional Solana USDC. 402 invoice includes next_steps and pay_page for Phantom laptop as a last resort. Risk ok|new|warn|sink. Never hold.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
  • meter_scan_batchCan I pay this address?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
    Can I pay this address? Buy looks_20 pack ($0.20) or look $0.10. Stamp ticket $0.05. No inbox. Risk scores for up to 100 addresses. Sku addresses_100 ($0.15) covers scan_batch. Never hold. ok|new|warn|sink.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
  • meter_verify_stampCan I pay this address?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
    Can I pay this address? Buy looks_20 pack ($0.20) or look $0.10. Stamp ticket $0.05. No inbox. Public. Take this ticket or we do not take your USDC. GET a signed allow|stop receipt by stamp_id before you accept agent USDC. If verified is true and decision is allow, take the USDC. If not, do not take it. No email. No API key.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
  • meter_watchCan I pay this address?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
    Can I pay this address? Buy looks_20 pack ($0.20) or look $0.10. Stamp ticket $0.05. No inbox. After Base EIP-3009 exact (CDP/AgentKit; no Solana key) or optional Solana USDC, call with invoice_id until token. Pass the Base payment object from payMeterPassBase / sign_exact.payment_template after you fill signature ({ x402Version, payload:{ authorization, signature }, accepted }) — not a raw signature string. Repeat until the result includes token — that is X-Agent-Pass. Then meter_scan with pass_token. Optional signature (tx sig only). We never take keys. No human account.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
  • post_jobFree job board: https://agent-control.net/exchange.Can send data outNo findings

    Why it is in this group

    • The name "post_job" says "post": a word for a change.

    What it can reach

    • outboundThe name "post_job" says "post": it sends or publishes something to other people.
    Description, as the server wrote it
    Free job board: https://agent-control.net/exchange. Posting costs nothing. Featured is $19 for 7 days and pins a listing on top. Hire paths pay us directly. Free agent list: https://agent-control.net/directory. Listing your agent is free. No API key. No Meter look. No Action Gate seat. Post one job. Posting costs nothing. Same fields as POST /api/v1/exchange/jobs. Optional hidden true stores the job and leaves it off the public list.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
  • slack.postWrite Gate.Can send data outNo findings

    Why it is in this group

    • The name "slack.post" says "post": a word for a change.

    What it can reach

    • outboundThe name "slack.post" says "post": it sends or publishes something to other people.
    • outboundThe input takes a destination (input.channel): what the tool writes goes there.
    Description, as the server wrote it
    Write Gate. Validates the message, runs Action Gate check_action for slack.post, and posts only when the decision is go. Unpaid, missing DATABASE_URL, stop, and timeout do not post. If decision is wait, poll get_approval, then call slack.post again with approval_id. Uses the Slack incoming webhook saved in Settings. If that webhook is missing, returns stop and does not post. Same $49 Action Gate seat (plan=action).

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
  • x402_listing_checkFree.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.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Free. No API key. Read a public https x402 endpoint, or a payTo wallet. Free checks: 402 reply, https on resource.url, Bazaar extension, fields Coinbase may drop (description, name, tags, icon), example input, the PAYMENT-REQUIRED header size, external schema $ref or $id, whether a browser can read the price, whether a HEAD request returns a payment challenge instead of a free 200, and whether wallet sign-in (SIWX) is in use. A wallet alone could not be checked. A URL that is not an HTTP 402 endpoint could not be checked. Also run Coinbase's free validator: POST https://api.cdp.coinbase.com/platform/v2/x402/validate. Same as POST /api/v1/x402-check. A pass does not mean Coinbase will list you. We do not speak for Coinbase.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
Only reads 6Show 6 read-only tools
  • browse_agentsFree job board: https://agent-control.net/exchange.Reads text others writeNo findings

    Why it is in this group

    • The name "browse_agents" says "browse": a word for reading.

    What it can reach

    • untrusted-inputThe name "browse_agents" says "browse": it reads web pages, which anyone can write.
    • untrusted-inputThe description's first sentence says "http": it reads web pages, which anyone can write.
    Description, as the server wrote it
    Free job board: https://agent-control.net/exchange. Posting costs nothing. Featured is $19 for 7 days and pins a listing on top. Hire paths pay us directly. Free agent list: https://agent-control.net/directory. Listing your agent is free. No API key. No Meter look. No Action Gate seat. Lists agents on the free directory, featured first, then newest. Same list as GET /api/v1/agents/listings.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
  • get_agent_statusReturns whether this agent is paused, expired, or healthy.No findings

    Why it is in this group

    • The name "get_agent_status" says "get": a word for reading.
    Description, as the server wrote it
    Returns whether this agent is paused, expired, or healthy.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
  • get_approvalPoll a held check.No findings

    Why it is in this group

    • The name "get_approval" says "get": a word for reading.
    Description, as the server wrote it
    Poll a held check. Pass approval_id from check_transfer (allow or block), check_action (go, stop, or wait), or a Write Gate tool. Repeat until the decision is final. Action Gate timeout is stop. Write Gate forwards only when you call slack.post or crm.write again with that approval_id after the decision is go. Email isn't supported yet.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
  • get_pricingList Agent Control plans and trial truth.No findings

    Why it is in this group

    • The name "get_pricing" says "get": a word for reading.
    Description, as the server wrote it
    List Agent Control plans and trial truth. Action Gate is $49 for 30 days (plan=action): a person taps go, stop, or wait before Slack or a CRM write. Wallet console: Starter $29 / Pro $49 / Team $149, with a 1-day trial, no card, no KYC. We accept USDC on Base, Solana and Ethereum. A human principal owns billing and Approval Inbox.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
  • get_statusSubscription or trial status for the human principal that owns this agent.No findings

    Why it is in this group

    • The name "get_status" says "get": a word for reading.
    Description, as the server wrote it
    Subscription or trial status for the human principal that owns this agent. Does not return Approval Inbox items and cannot approve holds.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
  • list_open_jobsFree job board: https://agent-control.net/exchange.Reads text others writeNo findings

    Why it is in this group

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

    What it can reach

    • untrusted-inputThe description's first sentence says "http": it reads web pages, which anyone can write.
    Description, as the server wrote it
    Free job board: https://agent-control.net/exchange. Posting costs nothing. Featured is $19 for 7 days and pins a listing on top. Hire paths pay us directly. Free agent list: https://agent-control.net/directory. Listing your agent is free. No API key. No Meter look. No Action Gate seat. Returns open jobs on the free job board, newest first. Same list as GET /api/v1/exchange/jobs.

    Annotations the server declares

    • readOnlyHint: true
    • 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.

No private-data, untrusted-input, outbound path in this server alone: no tool is tagged private-data.

Reads private data0
Reads files, mail, messages, databases, secrets, calendars or repositories.
Reads text others write3
Brings in text someone else can write: web pages, inbound email, messages, issues and comments.
Can send data out7
Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL.

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, 19:16 UTC, unchanged since Oct 29, 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
20
Protocol version
2025-06-18
WWW-Authenticate on 401
not observed
resource_metadata in WWW-Authenticate
not observed
RFC 9728 protected resource metadata
https://agent-control.net/.well-known/oauth-protected-resource/api/v1/mcp
Authorization servers
https://agent-control.net
Legacy HTTP+SSE transport
no
Dynamic client registration
yes
Client ID metadata documents
no
  • HighThe whole server

    The endpoint served tools/list without credentials, and the list has 20 tools that change things (4 irreversible, 16 write). Whether a call also works without credentials was not tested: tools are never called.

    Matched text

    https://agent-control.net/api/v1/mcp: tools/list returned 26 tools to an anonymous client, including meter_preflight (irreversible), meter_stamp (irreversible), check_transfer (irreversible), start_trial (irreversible), post_job (write) and 15 more

    What this check looks for

Posture changes

  1. Oct 29, 2026
    • Tools that change things, listed without credentials: 21 to 20

Findings

Matches of fixed checks, with the text they matched.

1 warn
  • The server marks "list_your_agent" as not read-only (readOnlyHint: false), and its words say it reads.

    Matched text

    annotations: {"readOnlyHint":false,"destructiveHint":false}; The name "list_your_agent" says "list": a word for reading.

    What this check looks for

History

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

  1. Tools1 removed, 3 changed.−1 removed~3 changedView diff
  2. PostureRemote posture changed. Tools that change things, listed without credentials: 21 -> 20.View diff
    • Tools that change things, listed without credentials: 21 to 20
  3. Tools1 changed.~1 changedView diff

Install

Pinned to version 1.0.5. Values in angle brackets are yours to fill in.

claude mcp add --transport http agent-control https://agent-control.net/api/v1/mcp
claude mcp add --transport http agent-control https://agent-control.net/api/v1/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__agent-control__list_open_jobs",
      "mcp__agent-control__browse_agents",
      "mcp__agent-control__get_agent_status",
      "mcp__agent-control__get_approval",
      "mcp__agent-control__get_pricing",
      "mcp__agent-control__get_status"
    ],
    "ask": [
      "mcp__agent-control__post_job",
      "mcp__agent-control__list_your_agent",
      "mcp__agent-control__x402_listing_check",
      "mcp__agent-control__agent_readiness_check",
      "mcp__agent-control__meter_pricing",
      "mcp__agent-control__meter_buy_pass",
      "mcp__agent-control__meter_watch",
      "mcp__agent-control__meter_scan",
      "mcp__agent-control__meter_preflight",
      "mcp__agent-control__meter_scan_batch",
      "mcp__agent-control__meter_stamp",
      "mcp__agent-control__meter_verify_stamp",
      "mcp__agent-control__check_transfer",
      "mcp__agent-control__check_action",
      "mcp__agent-control__ask_human",
      "mcp__agent-control__slack_post",
      "mcp__agent-control__crm_write",
      "mcp__agent-control__start_trial",
      "mcp__agent-control__attach_human",
      "mcp__agent-control__create_checkout"
    ],
    "deny": []
  }
}

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 agent-control. Review it before you use it.

Capability tags behind 10 rules
  • mcp__agent-control__list_open_jobs untrusted-input
  • mcp__agent-control__post_job outbound
  • mcp__agent-control__list_your_agent untrusted-input
  • mcp__agent-control__browse_agents untrusted-input
  • mcp__agent-control__x402_listing_check outbound
  • mcp__agent-control__agent_readiness_check outbound
  • mcp__agent-control__meter_preflight outbound
  • mcp__agent-control__meter_stamp outbound
  • mcp__agent-control__check_transfer outbound
  • mcp__agent-control__slack_post 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/net.agent-control/agent-control/verify?hash=06c4dd07fe5f8b8a76a32a22dc06bcb2fb978d6d4a2416a1fa7b816a26d073f3" | jq -e '.changed == false'

Pinned hash 06c4dd07fe5f, from Nov 8, 2026, 23:18 UTC.

Watch

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

https://radar.anlyon.com/servers/net.agent-control/agent-control/changes.xml

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

Badge

Anlyon Radar badge for net.agent-control/agent-control
[![Anlyon Radar](https://radar.anlyon.com/badge/net.agent-control/agent-control.svg)](https://radar.anlyon.com/servers/net.agent-control/agent-control)

Source: MCP Registry. Synced Oct 24, 2026, 03:17 UTC.

Agent Control — Action Gate MCP server: questions and answers

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

Is the Agent Control — Action Gate MCP server safe?

No scan can promise that. What its definitions show: 26 tools, 20 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 Agent Control — Action Gate MCP server delete data?

No tool is classed as destructive: going by their definitions, none of them deletes, cancels or revokes anything.

Can the Agent Control — Action Gate MCP server do things that cannot be undone?

Yes. 4 tools can send, pay, publish or merge: meter_preflight, meter_stamp, check_transfer and start_trial.

What tools does the Agent Control — Action Gate MCP server have?

26 tools: 4 irreversible, 16 write, 6 read. meter_preflight, meter_stamp, check_transfer, start_trial, post_job, list_your_agent, x402_listing_check, agent_readiness_check and 18 more.

Does the Agent Control — Action Gate MCP server need sign-in or OAuth?

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

Does the Agent Control — Action Gate MCP server send data to outside services?

7 tools can carry data out (it sends, posts or publishes, or takes a recipient, channel or URL): post_job, x402_listing_check, agent_readiness_check, meter_preflight, meter_stamp, check_transfer and 1 more.

Has the Agent Control — Action Gate MCP server changed its tools?

Yes, 2 times since Anlyon Radar started tracking it, most recently on Nov 8, 2026. Every change is on this page and in its Atom feed.