285 Free Prompts to Save Time
Your inbox isn't the problem. Neither are your meetings. It's the hours you spend on work AI could do in minutes.
285 copy-paste prompts: built for email, meetings, admin, and delegation. Each one has a clear role, a specific task, and a defined output.
No setup. No prompt engineering. Just copy, paste, and get your time back. Free when you subscribe to Mindstream.
The week's theme is agents stepping into the real world and the guardrails going up around them. Walmart opened its store to AI shopping agents with spending limits, Waymo lined up money to grow its robotaxi fleet fivefold, and Tesla started building its first production robots as profit fell. We checked the retailer's own terms, the fleet numbers and Tesla's shareholder letter before writing, and the caveats are in the copy.
TODAY IN AI
3 things that happened while you were busy
1. Walmart is letting AI agents shop for you, with limits you set.
Walmart launched an Agent Gateway that lets approved AI agents search its inventory and buy for you across Walmart and Sam's Club, and Agent Profiles roll out next month so each agent's purchases show up separately. You can require approval on every order or pre-authorize up to a spending limit, agents never get your card details, and you can delete an agent's profile any time. Read TechCrunch's report. Walmart calls it the "opposite approach" to Amazon, which has started blocking third-party agents and pushes its own assistant instead.
2. Waymo lined up $5B to grow its robotaxi fleet fivefold.
Waymo plans to take its roughly 4,000-car fleet to about 8,000 by year end and 20,000 by the end of 2027, funded by a $5B loan, its first debt deal, from PIMCO, Blackstone and Sixth Street, according to Semafor. The company runs paid rides in 15 US cities and aims to carry its first riders in the Madrid region in 2027. Read TNW's report. Waymo declined to comment, so treat the exact numbers as reported rather than confirmed.
3. Tesla started building Optimus robots as quarterly profit fell 43%.
Tesla said in its Q3 shareholder letter that it has begun making Optimus humanoid robots at its Fremont plant at a "production-quality level" for the first time. Net profit fell 43% to $787 million from $1.37 billion a year earlier, on revenue of about $28B, which Tesla linked to heavier spending on Optimus, AI and autonomy plus lower regulatory credit sales. Read TechCrunch's report. Tesla has not said how many robots it is making or when they go on sale.
Sponsored by
Finally, a Monday that doesn’t feel like one.
Imagine a Saturday where you're not catching up on the CRM you didn't touch all week.
HubSpot's new self-updating CRM doesn't wait for you to log back in. Friday's calls are already logged. The deal that needs a nudge Monday morning is already flagged. The automation you set up once keeps running without you thinking about it. None of it needed you over the weekend.
Customers are already seeing it pay off: 45% more deals closed. Finally, you start Monday caught up, not behind.
TODAY'S TUTORIAL
Let an AI agent shop for you without losing control
Shopping agents like Meta's Muse and Claude Duet can now buy things on your behalf, and retailers are building the controls to keep that safe. Six banks, including Bank of America and NatWest, warned in September that agents can enter card details into sites or steer you toward weaker payment protections, per The Neuron. Here is how to use one without handing over the keys. See the controls Walmart built.
1. Give the agent its own profile. Use a store's official agent connection, like Walmart's Agent Gateway, so the agent's purchases stay separate from yours and you can switch it off without touching your main account.
2. Set a spending limit. Treat it like adding an authorized user to a card: cap how much the agent can spend before it has to ask you, and start the cap low.
3. Require approval above a threshold. Let it reorder small staples on its own, but make anything over a set amount wait for your yes. Most agents and stores now support both modes.
4. Never paste your card into the agent. A real integration keeps payment on the store's side. If an agent asks you to type card numbers directly into a chat or a site it opened, stop.
5. Know the returns path. Before you let it buy, check that the store's normal returns apply when an agent orders the wrong thing, and keep your own order confirmations.
Good to know: Amazon is going the other way and blocking outside agents, so an agent that works at Walmart may not work there. Match the agent to the stores that actually support it.
FROM THE FRONTIER
The agents are moving into the physical world, and everyone is writing rules at once
The storefront. Walmart and Amazon just split on the biggest question in retail: whether to let AI agents buy for customers. Walmart built a gateway with spending caps and per-agent profiles; Amazon is blocking third-party agents and asked Perplexity's Comet to stop offering Amazon shopping, citing customer-service problems, per The Neuron. See Walmart's terms.
The road. Waymo's plan to reach 20,000 cars moves robotaxis from pilot to infrastructure, and it is borrowing money to do it rather than waiting on profit. Scale is now the strategy, which also means more cars on more streets before the rules fully catch up. See the fleet targets.
The factory floor. Tesla building Optimus units while profit drops shows how much the big players will spend to get robots into the world early. Unitree, meanwhile, posted a demo of a humanoid disarming an attacker and lifting a car, though it has not said whether the actions were autonomous or choreographed, or shared any specs. See the Unitree demo coverage.
The pushback. Guardrails are arriving from courts too. Two Chicago residents filed class actions against delivery-robot firms Coco and Serve, alleging the robots scanned faces and hands without consent and sold the data for AI training; both companies deny it, per Superhuman. Read the report.
The reframe. For years AI lived behind a screen, where a mistake meant a bad answer. Now it drives, shops and lifts, where a mistake means a charge, a crash or a collision. The companies racing to deploy and the banks, judges and retailers writing limits are really arguing about the same thing: who is liable when the agent gets it wrong.
MONEY FORWARD
Three ways to turn this week's AI news into income
1. Make store listings agent-ready. As agents start picking products, clean titles, clear specs and structured data decide which items get chosen. Offer small online sellers a listing cleanup so their products surface when an AI shops, the same way SEO once decided search. See where agent shopping is heading.
2. Safe-agent setup for busy people. The tutorial above is an hour most people will not spend themselves. Sell it as a fixed-price session: connect the agent, set spending caps and approvals, and leave a one-page rulebook of what it may buy. See the controls to configure.
3. Meeting notes with a free tool. TwinMind transcribes meetings for free and builds a searchable record. Offer local professionals a notes-and-summary service: sit in on their calls, deliver clean action items the same day, and charge per meeting. See TwinMind.
GET THE EARNINGS GUIDE
Want the step-by-step version of ideas like these? Get The Aiwire earnings guide at theaiwire.co. These are ideas, not income promises, and results depend on your skills, market and effort.
IN THE KNOW
What people are actually watching and sharing
Legal AI's big round. Legora raised $550 million at an $8.8 billion valuation to expand its legal AI business, per The Neuron. See The Neuron's roundup.
300,000 batteries, zero prototypes. GM said it tested more than 300,000 battery designs in computer models before building a single physical prototype, per The Neuron. See the item.
The AI paint boom. PPG expects its data center paint sales to triple this year, driven by demand for coatings that resist fire, rust and static, a quiet sign of how physical the AI buildout is. Read more.
A $3.6 trillion merger idea. Elon Musk is reportedly exploring a Tesla and SpaceX merger that could create a roughly $3.6 trillion company spanning rockets, AI and robots, per Euronews. Read the report.
A $1.45B robot brain. Eka Robotics raised $150 million to build AI that combines vision and force feedback for faster, more dexterous industrial robots, per TechCrunch via Superhuman. Read TechCrunch's report.
TRENDING TOOLS
5 tools worth a look this week
1. Loom Record for agent. Record a screen walkthrough and Loom turns your narration and clicks into a structured prompt for an AI agent. In open beta on Loom's Business and AI and Enterprise plans, with recordings capped at five minutes. See the docs.
2. Flint. Recursive's first model, a 40B-parameter coding model built to run locally; it does over 125 tokens per second on an M4 Pro MacBook and wants about 48GB of RAM. Weights are free on Hugging Face, though the page did not state a license. Read the details.
3. TwinMind. Records and transcribes meetings and lectures into a searchable "second brain" on iPhone, iPad, Watch and Mac. Free with unlimited transcription; Pro is $19.99 a month for advanced features. See TwinMind.
4. ChatGPT Locked Chats. Lock individual conversations behind a PIN or Face ID so they stay out of your history and search, rolling out on mobile. Handy if you share a device or run sensitive work in ChatGPT. Read the details.
5. Notion Auto. Notion's model router sends quick tasks to cheaper models and hard ones to stronger ones, which Notion says cuts average cost about 90% against Opus 5.5 in its own tests. Useful as agent usage, and its bills, add up. Read Notion's page.
PROMPT STATION
Write the house rules for your shopping agent
This prompt works in Claude or ChatGPT and turns "I guess it can buy stuff" into a clear policy you can hand any shopping agent. Do it before you connect one, so the limits exist in writing and not just in your head.
COPY AND PASTE THIS PROMPT
You are helping me write a clear policy for an AI shopping agent. My situation: [WHO IT SHOPS FOR, e.g. my household of four], a monthly budget of [AMOUNT], and stores I use: [LIST]. Write a one-page rulebook covering: 1) what it can buy on its own and the per-item and monthly limits, 2) what always needs my approval, 3) categories it must never buy, 4) what to do when an item is out of stock or the price jumped, 5) how it should report what it bought and how often. Write it as plain rules I could paste into the agent's instructions, and flag any rule that depends on a feature a given store might not support.
Tip: Swap [WHO IT SHOPS FOR] for a small office or a parent you help, and set tight limits first; you can always loosen them. For a sharper rulebook, add [my three past problem purchases] and ask the agent to write rules that would have caught each one.