1Password for Agents: AI Logs In, Never Sees Client Passwords

1Password for Agents lets an AI agent log in to a website for you without ever seeing the password. The agent asks, you approve in 1Password, and 1Password types the login straight into the site.

If you've ever been tempted to paste a client's password into an AI chat so it can "just finish the task," this is the safer way. Here's how it works, who can use it today, and what I'd tell a client.

The Problem It Solves

An agent is an AI that does tasks on its own, like filling in a form or checking an order status, instead of only answering questions. Agents are getting good at this. But the moment one hits a login page, it stops.

1Password's announcement on October 22, 2026, describes the two bad options you have today. You take over and type the password in yourself every time, which is slow. Or you paste the password into the prompt, which gives the agent your login "potentially beyond the task at hand."

The post also points out that agents can't use passwords saved in your browser or written in notes. So without something like this, you keep getting pulled back in.

How 1Password for Agents Works

When the agent reaches a login, it asks for the credential it needs. You get the request in 1Password, check which login it wants, and approve it. Then 1Password fills in the username and password directly on the site. The agent never gets the password or the one-time code.

In 1Password's own example with Muse, you pick which login the agent can use, and it doesn't get anything else in your vault. You can also disconnect the integration at any time.

There's a nice side effect for anyone who changes passwords often. Your logins stay in one vault, so when a password changes, you update it once and every connected agent uses the new one.

Which AI Agents Work With It

Not all at once, and this is where I'd be careful with clients.

Two integrations are "rolling out over the coming days": Muse from Meta, and Grok Bot from SpaceXSI. Anthropic's Claude Duet, OpenAI's ChatGPT Work and dots, and Townie from Town are "coming soon." A longer list, including Perplexity, Browserbase, Browser Use and Ramp, is also marked "coming soon."

1Password adds that "the safeguards and controls vary by integration." So check each agent's own setup before you rely on it.

Is It Free?

For many people, yes. 1Password says it's available "at no additional cost" for Individual and Family accounts on 1Password.com.

Two limits. Accounts on 1Password.eu or 1Password.ca aren't currently supported. And the post only mentions Individual and Family accounts, so don't promise it to a client whose team uses a 1Password Business account.

What This Means for VAs

More clients are going to want an AI agent set up for them. Logins are the scary part of that job, and now there's a clear rule you can offer: the agent asks, a human approves, and nobody pastes a password anywhere.

Here's how I'd put it in practice. Write one line into your client SOP (standard operating procedure, your written how-we-do-things doc): "AI agents get logins through 1Password approval only, never a pasted password." Then list which agent the client uses and whether it's live yet or still "coming soon."

If your client is waiting on Claude Duet or ChatGPT Work, you have time to tidy their vault first. Remove old logins and name items clearly, so the right one is easy to spot when an approval request comes in.

You become the person who makes the AI safe to use. That's a skill clients will pay for.

If you're setting up agents for clients, tell me which one they picked.

Can Claude Make YouTube Videos? I Made a 4-Minute One

Yes, it can. This week I made a full 4-minute YouTube video with Claude: the research, the script, the voice, every scene, and the thumbnail. But it didn't happen the way most people imagine, and the part that surprised me most wasn't the video at all.

Here's exactly what Claude did, what I did, and what I'd tell any VA who wants to try it.

First, What "Making a Video" Meant Here

Claude didn't press a magic "make video" button. There's no camera and no AI-generated footage in this video.

I used Claude Code, the version of Claude that can run tools on your computer, not just chat. It wrote the video as a set of animated web pages, then recorded them frame by frame into an MP4. Think of it like building slides that move, then filming them.

The video is for my small test channel, The 1% Mindset. The topic: a new AI model that guesses an answer 93% of the time when it doesn't know.

How It Went, Step by Step

Research. I asked Claude to dig into the topic. It sent several research helpers out at once, then a separate set of fact-checkers who opened every source again. They checked 194 claims. Four were wrong, and none of those made it into the script.

Script. Claude wrote a 4-minute script in plain English, then checked its own opening against a retention checklist and rewrote the first lines. I read it and signed it off before anything else happened. That's the same job I'd do for a human writer.

Voice. The voice-over comes from Kokoro, a free voice model that runs on my laptop. Claude split the script into sections, voiced each one, then transcribed the audio back to check that every word came out right.

Visuals. Every chart and post you see is a real screenshot of the source. Claude added the red hand-drawn circles and underlines, timed to the voice. The only made-up art is the cartoon test sheets and bunny from the thumbnail, plus a few simple bar graphics.

Thumbnail. Claude wrote the prompt for Nano Banana 2 (Google's image model), then added the "100% SURE" text on top.

The final file: 4 minutes 12 seconds, full HD.

What Surprised Me

The checking. After the render, Claude pulled a frame from every 2 seconds of the video and looked at all of them, like a very patient editor.

It caught real mistakes. One red circle sat on the wrong bar of a chart, because the website had shifted its layout after the screenshot. Another section sat empty for 8 seconds. Both got fixed before I ever saw the video.

On a second video, it caught something worse: a screenshot that showed numbers we hadn't been able to verify. It cropped them out. As someone whose job is getting details right for a client, that's the part that won me over.

Is It Free?

Not completely. Claude Code needs a paid Claude plan. The voice model is free, and the thumbnail image costs a little through Google.

Time is the bigger cost. This took most of a day, with me reviewing at every step. It's still a fraction of what one person would need to research, write, voice, animate, and edit the same video by hand.

What This Means for VAs

If you manage YouTube for a client, this changes what you can offer. You don't need to be a video editor to turn a script and real sources into a clean explainer, a Short, or a slide-style video.

But notice who made the decisions. Claude did the heavy lifting, while I picked the topic, approved the script, chose the look, and said no when something felt off. That's still VA work. It's just a bigger version of it.

Start small. Ask Claude to turn one of your client's blog posts into a 30-second Short script, and see how much of the rest you'd want to hand over.

If you try it, tell me how it goes on the comments.