Short answerTunnel gives your business an address AI agents can message — over Model Context Protocol or plain HTTP — and relays what they send to an inbox in your dashboard. You reply there and the agent reads it on the same conversation. The relay never hands over your phone number or your email: messaging you is not a way to obtain them. Only registered agents can write, under per-agent quotas and burst limits, with an alarm on anyone messaging many merchants at once.
Watching what assistants say about your business is a read. Anyone outside your business can do it, without asking you.
Being named in an answer is half of it. The other half is what happens when someone wants to actually buy — and an assistant has no way to reach you.
Any assistant or custom agent can open a conversation with your business. No integration on your side.
Every agent conversation lands in the same place. You reply like you would to an email.
Per-agent quotas, burst limits and a fan-out alarm stop one bot farming a hundred merchants.
Your inbox shows a message. This is the request that produced it — one tool call, from a registered agent, over Model Context Protocol.
https://api.tunnelpowered.com/api/mcpexample{
"jsonrpc": "2.0",
"id": 1,
"method": "tools/call",
"params": {
"name": "contact_business",
"arguments": {
"slug": "brutaria-de-pe-colt",
"agent_name": "A shopping agent, on behalf of a customer",
"subject": "30 portions delivered Friday evening",
"message": "Can you deliver 30 portions to Botanica on Friday at 19:00?"
}
}
}An excerpt. What the agent gets back is a conversation id and a secret token: those two together are the only way it can read your reply, and the token cannot be reissued if it is lost.
Five steps. Four of them are ours; the last one is a person reading a message.
Through the knowledge-base search, the MCP server, or a direct link. Your record carries the messaging address as a field, so the agent does not have to guess.
One MCP tool call, or one HTTP POST. No integration, no SDK and no negotiation on your side — the endpoint is the same for every business on the platform.
The agent must be registered and hold a token. We check its quota, its burst rate, and how many other merchants it has contacted recently before anything reaches you.
Every agent conversation appears in one place in your dashboard, threaded, with the agent identified.
You answer the way you would an email. The reply goes back on the same conversation, so the agent can carry it into the answer it is writing for its user.
Opening an inbox to anything that can make an HTTP request is only a good idea if the abuse case was designed first.
Reading is open to anyone. Writing requires a registered agent identity and a token, so every message has something attached to it that can be rate-limited and revoked.
A single agent cannot flood one merchant, and cannot make its volume look reasonable by spreading it thin over time.
One agent contacting many merchants in a short window is the signature of harvesting rather than buying. It trips an alarm regardless of whether each individual merchant is under quota.
The relay address is ours. Your phone number and email are not in the record an agent reads and are not revealed by messaging you.
It does not place orders or take payment. It is a conversation channel; the transaction happens the way it already does.
It does not answer for you. There is no bot writing replies in your voice, and we will not add one without saying so loudly.
It does not oblige any assistant to use it. Consumer ChatGPT will not spontaneously message your bakery today. Most agent traffic right now is developer-built, and this is infrastructure for when that changes — cheap to have and impossible to retrofit in a hurry.
It does not guarantee the message is genuine. A registered agent is an accountable one, not a vetted one. Treat an agent enquiry the way you would treat an email from a new customer.
Priced by reach: how many businesses can be messaged through it, and whether a person has verified that they are yours. The address, the quotas and the human routing are the same on every tier — a machine that cannot reach you is not a cheaper product, it is a broken one.
One business, reachable by any registered agent, with a person at your end.
For several locations, and for an agent that needs to know a person checked.
More sites, more profiles, a higher crawl or probe cadence, or an arrangement that does not fit a card. Negotiated per business, invoiced rather than charged to a card.
Endpoints, tool schemas, authentication, rate limits and the dispute process are documented for the developer building the agent rather than for the merchant receiving the message.
Read the agent documentationAny of them, in principle: the MCP server and the REST endpoint are open to any registered agent, and there is nothing vendor-specific about either. In practice, whether a given consumer assistant chooses to use it is that vendor’s decision, not ours. Today the realistic caller is a custom or developer-built agent.
Nothing beyond what your listing already needed. If you have no website there is nothing to integrate at all — the messaging address exists as soon as your profile is published. If your listing is built from a website, it comes with the same one line that makes the rest of that listing live: an agent can reach a site we have verified as integrated and cannot reach one we have not. Either way you read and answer messages in the dashboard inbox, and there is no SDK, no callback and no endpoint for you to host.
Only if you publish it. Your phone number is a field on your own record — fill it in and it goes out as `telephone` in your structured data, because a customer asking an assistant how to call you should get an answer. Leave it out and nothing else puts it there. What the relay never does is hand it over: an agent messages an address of ours, and replying does not reveal your number, your email, or anything else you did not choose to publish.
Three things: writing requires a registered agent identity, each agent has a quota and a burst limit, and contacting a large number of merchants in a short window trips a fan-out alarm even when each individual merchant is within quota.
Nothing bad happens to your record — an unanswered conversation is not a penalty. But the agent gets no answer, and whatever it was asking about goes elsewhere. The inbox exists to be read.
Because it runs in the other direction. A tool that measures visibility reads public sources and reports what it found; nothing about doing that gives an agent somewhere to send a message, or gives you somewhere to read one. An inbox needs an address that resolves to your business, an identity check on whoever writes to it, quotas that hold up against a bad actor, and a person who answers. That is infrastructure rather than a report, and it has to already exist on the day the first agent tries.
Yes. They are separate products with separate plans, and an account can hold one, the other or both. A Landline-only account gets the business profile, the agent address and the human routing, and none of the crawling or visibility measurement — which is also why it is not charged for them.
Because there is nothing in it yet. Everything a Business tier could add today — deeper crawls, more keywords, more article drafts — belongs to AI GEO, and the two things Landline actually meters (how many businesses, and whether a person verified them) are already at their ceiling on Pro. We would rather show two tiers that differ than three where two are the same price for the same thing.
The account drops to the free tier and the data stays. Your record, your profiles and your conversation history are intact; what changes is the crawl depth, the check cadence and whether a person will verify a new profile. Upgrading again restores the rest immediately.
Per account. A plan covers the account and the allowance is a number of businesses inside it — one on free, three on either paid tier. Beyond three, the extra profiles are arranged with us.
We ask four assistants what they say about your business, and send you the answers they gave — including the parts that are wrong. Free, and yours to keep.
By submitting, you agree to our Terms of Service and Privacy Policy