user avatar

Sailor: the money agent

I’ve spent the last three years obsessed on building one idea: putting a money agent in every pocket. An agent that doesn’t just give you the best financial knowledge but an agent that also transacts safely on your behalf and handles anything money-related.

I’m doing this because I started my career in the financial advisory and asset management industries and know from the inside how biased, conflicted, slow, and expensive they are. Then I went deep into crypto and later into AI, and realized they both together unlock a product that will change how humans interact with money forever. They essentially fix what I experienced.

So this article is about the plan we’re executing to build that product: Sailor. What it is today, and the three stages that take it from an open source harness for builders to the best money agent in the world.

Sailor today: harness for DeFi agents

Three weeks ago we launched the first version of Sailor as an open source harness that turns any coding agent into a builder of DeFi agents.

The coolest features of a Sailor DeFi agent are:

No access to your private keys

Runs locally on your computer

Covers all DeFi use cases: trading (spot, perps), yield (lending, borrowing, looping, liquidity providing), prediction markets, payments, everything

To give it a try, you just have to paste the following package into your coding agent of choice:

The package contains everything you and your coding agent need: CLI with one command for every step, open source SDK, more than 20 skills, a local dashboard, and a journey split into stations that takes you from zero to a running DeFi agent optimizing token expense.

Sailor runs on Sail Protocol, which is decentralized and permissionless infra that makes Sailor Agents non-custodial and safe by default. I explained it in depth last week, take a look:

Sailors are already running DeFi agents for a vast amount of use cases, including portfolio managers, traders, autonomous investors, yield optimizors, and more. Here are a few examples:

Based on feedback, we’re shipping two new features really soon: Shipyard, a simulation sandbox where your agent can trade simulated funds against real market conditions before touching real funds, and Harbor, a library of pre-built agents you can run in minutes while expending a fraction in tokens.

Sailor tomorrow: the DeFi agent

The honest limitation of Sailor today is that it is a L2 harness. Sailor runs on top of another harness (your coding agent), which comes with limitations in user experience, an entry barrier for non-builders, and blockers on the next features we wanna ship.

Stage two means turning Sailor into an L1 harness so it becomes its own agent. This includes shipping the following:

Desktop app. The whole experience redesigned for non-builders, while keeping Sailor available in the the terminal too.

LLM orchestrator. Sailor runs the models itself, picking the right one for each task: research, strategy design, operation, etc.

Long-term memory. An agent that knows you. Your portfolio, your history, your risk preferences, your goals.

Top-tier onchain data and intelligence. Ask anything about the market and get answers grounded in live onchain data.

Cloud automation. We believe local is the future but cloud automation is key for onboarding.

This is the stage where we will have a product strong enough to start getting escape velocity and PMF with crypto-natives.

Sailor after tomorrow: the money agent

This is the product my mom will use so let me describe it:

Knows you better than any private banker ever could. Your portfolio, your income, your goals, your habits, your risk, your preferences. And unlike a banker, it’s present for every single decision you make.

Your data stays yours. The agent’s memory belongs to you and not to the model. Your financial life is never shared with them and never becomes training data.

It’s portable, lives everywhere. Your phone, your laptop, your browser, new agentic devices. Cloud when you want convenience, local when you want full sovereignty. One agent with one single memory wherever you are.

It’s connected to all forms of money and actions. Onchain and TradFi with ramps, market data, payments, cards. Paying, saving, trading, moving, borrowing, investing.

In terms of distribution, I don’t only see it as a retail product but also as a potential whitelabel solution. Any financial advisor or asset manager will be able to ship their own money agent on top of Sail with their brand, their strategies, and UX. On our rails.

So the destination is: everyone, everywhere, with a personal money agent that knows you, informs you, and executes safely for you.

Install it today:

And If you have any questions and/or feedback, please head to Sail’s Discord Server.

Source link

Share:

Leave a Reply

3 latest news
News Archives
On Key

Related Posts

user avatar

Sailor: the money agent

I’ve spent the last three years obsessed on building one idea: putting a money agent in every pocket. An agent that doesn’t just give you

curated design references for AI agents

curated design references for AI agents

Reaching 1,146 sites, 3,205 captured sections, 2,811 full-page captures, 568 fonts and 507 creators. search_websites Full-text search across site names, descriptions, and design tags. browse_websites

No, local models will not win

No, local models will not win

Every time a new open-weight AI model is released, people say that local models are the future. Why spend billions of dollars building out datacenters