Building Assets: Easy's One-Person Business Methodology

A course can keep selling after the teaching is recorded. A software tool can serve another customer without another day of custom development. That distinction between selling hours and maintaining an asset runs through The One-Person Business Methodology (《一人企业方法论》), by the independent developer Easy (easychen). The book grew from a popular GitHub essay into a book of nearly 60,000 Chinese characters, published openly under a Creative Commons license (CC BY-NC-SA). In 2026 it gained something new: a set of agent skills that walk a founder through the method step by step.

Easy writes from experience. He built Server酱, a simple push-notification service with more than 200,000 users, and other independent tools, yet he says frankly that his software never covered his living costs. What did was a by-product: courses he made while building the software. That story, an independent developer whose side output turned out to be the business, runs through the whole book.

Below are the ideas we find most useful, with our notes on what AI agents change. The book's own wording is in Chinese; the translations are ours.

What a one-person business is, and isn't

Easy defines a one-person business as "a business led by an individual or a personal brand." Three clarifications follow:

It doesn't have to be a registered company. It can be a side business. It can also have a small team, as long as one person leads it. He later switched the book's title from 公司 (company) to 企业 (enterprise) to avoid confusion with the legal single-shareholder company.
It is not a freelancer selling time. It earns from assets and uses leverage to scale.
It is not a startup. It does not need investors, so it is allowed to be "small and beautiful," better rather than only bigger. It can become a startup later, on the founder's schedule.

Selling time cannot scale

The book's central argument is arithmetic. Even at a strong freelance day rate and ten-hour days, one person's income from selling hours has a hard ceiling. So the founder needs income that does not depend on hours, and Easy borrows Robert Kiyosaki's definition of an asset: something that puts money in your pocket while you are not working.

There are two ways to get assets: buy them if you are rich, or create them if you are not. Most of us create. The book's 2.0 version adds a systems view on top: a one-person enterprise is not one business but a pool of assets, such as a tool, a course, a newsletter or a community, that the founder maintains and reallocates over time. Hence a chapter titled "One-person enterprise ≠ one-person business."

What agents change: they can reduce the effort of drafting content and building software. The time saved depends on the work and the review it needs. Lower production effort also makes it easier to create more assets than a founder can maintain. The book's discipline of treating each asset as something that must keep "putting money in your pocket" matters more, not less.

Why a small player can win

Easy's explanation is the shifting boundary (边界变动). Incumbents are optimized for the river as it used to run. When technology, policy or habits change, a new branch opens and old moats can be bypassed. His example is WordPress and its ecosystem of more than 60,000 plugins: once nearly impossible to match, but if AI coding matures, rewriting those plugins for a new platform might take a month.

Scale, in turn, comes from what Naval Ravikant calls new leverage: products with zero marginal cost of copying, such as code, content and media, which need nobody's permission. And from new infrastructure: outsourcing, crowdsourcing and renting instead of building, which replaces what used to need hired staff or capital.

What agents change: AI agents are both a boundary shift and a new kind of leverage. Agents also turn part of the labor that used to need outsourcing into software, which is exactly what the one-person biotech Mithril did with its coordination work.

Choose the track: strong needs of a small group

The book's rule for choosing a market is to serve "the strong needs of a small group, not the weak needs of the masses." A small group is cheaper to reach through content, search and what Easy calls side-project marketing: a free tool built for one niche that brings in its users.

Then avoid competing at all. The non-competition strategy has two moves: become part of an ecosystem (his examples are businesses built on Notion and Webflow, such as tools that turn Notion pages into websites), and increase uniqueness, through work that is not standardized or products people collect rather than replace.

Where competition can't be avoided, use structural advantages that a larger company cannot copy:

By-product advantage: sell what your main work already produces, such as methods, workflows, templates and industry knowledge, at almost no extra cost.
Third-party advantage: do what the official platform won't or can't, such as aggregating across competitors.
Low-cost advantage: once you reach your minimum viable profit, enough income to live on, you can keep going for years. Funded startups cannot afford that patience.
Dedicated-service advantage: with few customers, each can get personal service, provided the price is high enough.

What agents change: repeatable by-products can become candidates for automation. Your workflows can now be packaged as agent skills, custom assistants or automated reports. Easy's book does exactly this with its own method.

Manage risk: start as a side project

Easy's first piece of advice is not to quit your job on impulse. Start as a side project, where your salary makes the business resilient while it is unprofitable. Test the core assumption first with a minimum viable product, a landing page or a pre-sale. He also borrows from Nassim Taleb's Antifragile: small, quick experiments, redundancy, keeping options open, and asymmetric bets where the downside is small and the upside is large.

Build infrastructure you control

The last part of the book is practical: a one-person business needs four capabilities.

A user pool and a way to reach it. Distinguish native users on platforms you control, such as your own site, list or app, from followers on third-party platforms that can limit or charge for your reach.
A content pool and automation. Publish on your own site first, then sync to platforms, and bring readers back home.
A product pool and payments.
Crowdsourcing, so many users can each contribute a few minutes. One example is checking an AI translation.

Two thinking tools tie it together: the one-person business canvas, for planning, and a monthly report that compares your monthly recurring revenue with your minimum viable profit and with the income you'd need to work freely.

The method, as agent skills

In April 2026 Easy added nine agent skills to the repository. They take a founder through the method in order: an audit of the founder's resources (experience, network, skills, channels, assets, time and money), then niche positioning, a value proposition, a business model, an MVP, a conversion loop from reach to purchase, turning repeatable outputs into assets, and a dashboard review with stop-loss rules. An orchestrator skill coordinates them, asking one question at a time and saving each step's output for the next.

The skills apply the book's own approach: they turn a written method into a procedure that can be reused with a founder's inputs.

Applying the method

The book offers a way to examine a proposed business before spending more time on it. Name the buyer, the need, the asset and the cost of maintaining it. Then test whether anyone will pay. Agents can help produce the asset, while those questions determine whether it is worth producing.

The full book and the skills are free to read on GitHub. If you want help applying the method with AI agents, from the resource audit to the first automated asset, contact us.