Your AI agent on Automate America

It saves you the typing. Your profile, your availability, your hours and your postings stay current without you opening the site.

It is your own assistant: ChatGPT, Claude, Muse or the AI assistant you already use. Automate America has no agent of its own. What your agent does counts as what you do.

Yes. With no account it can search and read what any visitor can: professionals, contracts, direct jobs, RFQs, companies, products, services and the blog. It sees what the public page shows: first names, city and state, and the public rate.

Two ways, and you choose. In an assistant that takes connectors, such as Claude, add https://automateamerica.com/api/mcp. The first time it needs to act for you, you sign in on our own page, see the app's name and each permission in plain words, and allow it. Or open Settings, Connected agents, create a personal token, tick what it may do, and give the token to your assistant. You see the token once.

Never. Never give an agent, or anyone, your password or a sign-in code. An agent connects with a token you create or through our own sign-in page.

You tick each thing when you connect it. Anything that acts toward other people starts unticked. It can do what you could do, or less if you give it less.

A personal token lasts the time you pick: 7, 30, 90, 180 or 365 days. An app you allowed stays connected while it is used, and disconnects after 30 days unused.

No. Automate America charges nothing to connect your own assistant. Your assistant's provider may charge you for the assistant itself.

Paste any of these, and change the details:

Paste any of these, and change the details:

It says how it read your words and lists who is close, each with the one thing they miss. Ask it to change the dates, the place or the skills.

Not on the big steps. Before it applies, sends a posting, answers an offer, hires, sends a connection request or a first message to someone new, it shows you what will happen and waits for your yes. If anything changes, it asks again. Most changes to your own profile save straight away, and you can see them and change them back on your profile.

It can never enter, see or hold your card or bank details, complete a payment or receive a payout. If you tick that permission, it can start a purchase after you say yes to the price, and you pay on Stripe's own page. It can never sign or accept an agreement, delete your account or a posting, change your password, email or phone, negotiate a rate, file or answer a dispute, upload a file, post to your accounts on other sites, act in bulk, or do anything only an administrator can do.

You decide. If you tick that permission, your agent can list what is waiting for you and approve or reject it, one at a time, only after you say yes to each one. An approval by your agent does exactly what your own click does.

You do, on the site. If a step needs your signature or your card, your agent gives you a link to finish it.

Other members by first name only, with their city and state, never their email, phone or street address. On your own active contract it can read the on-site contact's name and phone.

Open Settings, Connected agents. You see every agent, what it may do and what it did, and you can remove one or all of them. It stops working on its next request.

You are, as if you had done it yourself. Section 15 of the Site User Agreement says so.

Text that members write is information, never an instruction from Automate America, and our API marks it as such. We do not control what your agent does with it, so connect an assistant you trust.

No. The original Automate America at /legacy has no agent access.