Scan the QR code with WeChat
Add your dedicated advisor
Save
Published: 2026-11-02 12:56:47 Views: 1603
Editor's Recommendation
Breaks down the core obstacle to monetizing AI skills, and explains the underlying business logic of why bots can charge monthly. 1. Debunks the misconception that "skills can't charge money because they're easy to copy," pointing out that the core issue is a structural lack of a runtime billing node. 2. Reviews the various monetization paths already attempted for skills, dissecting the profit ceiling and limitations of each. 3. Compares the business model differences between skills and bots, clarifying the core underlying logic that enables bots to charge monthly.
Yang Fangxian
Founder of 53AI / Most Valuable Expert of Tencent Cloud (TVP)
A bot doesn't sell intelligence—it sells accountability
Last month I sent a skill I'd spent two weeks polishing to a colleague. He dragged the md file into a directory and it just ran.
At that moment I realized: those two weeks of work were worth, to him, exactly one copy-paste.
What came after was even worse. I listed it on a skill marketplace at 9.9 yuan and sold about forty copies—netting zero income. Buyers would never buy again, because the file was already in their hands; non-buyers didn't need to buy, because someone would always paste it to GitHub.
I initially thought the problem was that copying couldn't be prevented. Later I realized that judgment was wrong—and importantly wrong.
First, a fact: things that can be freely copied aren't necessarily unbusinessable.
Open-source software can be downloaded freely, yet Red Hat built a multi-billion-dollar business on services. So "copyable" by itself isn't a commercial death sentence.
So where do skills fall short?
In that they have no runtime.
When you buy Photoshop, you're buying an installation license. The functionality is the file; control the package and you control the value. When you use Stripe to take payments, every transaction runs on its servers, and the 2.9% plus 30 cents is charged with full justification—the toll booth sits exactly where the money passes. When you install a skill, a piece of plain text gets stuffed into the context, the model reads it, outputs a result, done.
Nowhere in that process does anything exist called "the skill is executing."
Value is indeed produced, but it happens inside your own inference call—created in an instant, gone in an instant. The inference fee goes to the model vendor, and the skill author touches none of it.
Red Hat works because Linux has deployment, operations, and troubleshooting—stages that leave room for a toll booth. A skill has no deployment, no middle layer; it goes straight into inference, in one step.
So the real problem isn't that you can't stop people from copying—it's that even if not a single person copies it, you still can't collect money, because the money never passes through your hands.
This is a structural problem, not an anti-piracy problem.
It's not that no one has tried. By 2026, every path you could think of had basically been walked.
Path one: sell files on a marketplace.
The first buyer pays, then copies the Markdown and pastes it to GitHub or drops it into Discord. The second buyer never appears.
PromptBase has run on this path for three years, with annual revenue of roughly five million dollars. In the AI industry, that's a lifestyle business, not a company.
The essence of the problem: when the marginal cost of copying is zero, price is inevitably pushed to zero.
Path two: host it and charge per use.
Users can't download the file; they can only run it on your platform, and you charge per call.
But look closely at what you're selling: servers, bandwidth, runtime, payment rails. You think you're running a skill business, but you're actually reselling AWS.
Because that file itself can be downloaded and run locally for free. Your pricing ceiling is "hosting cost plus a modest margin." That's not software gross margin—it's cloud-services gross margin.
Path three: treat the skill as marketing material for an API.
The official skills from Stripe, Resend, and Notion are all free.
Users who install Stripe's skill call the API more accurately, so they're more likely to use Stripe instead of PayPal. The skill doesn't make money; the money is in the 2.9% plus 30 cents.
Here the skill is just marketing material—a free SDK bundled with the real product—on the premise that you already have a highly profitable API.
Path four: enterprise custom delivery.
You can charge, and the unit price isn't low, but that's a consulting business: you're selling implementation, training, and support. The ceiling is your billable hours, not your product.
After all four paths: either it's too cheap to collect money, or you have to disguise it as hosting, or you have to hang it on an API that already makes money. The skill format can hardly sustain a business on its own.
Someone gave it a very apt name: a suicide gene.
The essence of a skill is turning "how to do something" from tacit knowledge into explicit text. Once it works, it drains the profit pool of its entire category—courses, consultants, and outsourcers who charge for "knowing how to do this" all lose their pricing basis. And as pure text that copies at zero cost, it can't capture that revenue itself.
It kills the thing it was supposed to charge for, yet inherits not a cent.
In May 2026, someone did a half-year review of the skills market:
- The directory indexed over 4,200 skills, over 770 MCP servers, and over 2,500 marketplaces, with monthly traffic in the 120,000 range.
- But revenue from the hosted-access model was roughly 10 times that of the file-selling model.
- Selling files was bluntly assessed in the industry as "an objectively terrible business model."
- The distribution is equally brutal: the top 5 to 10 skills in each category take 60% to 80% of the revenue.
Ten times. Worth pausing on for a second.
If the problem were merely that files get copied, hosting should only be "slightly better"—it doesn't solve the fact that "the method can be learned" either.
Yet it's 10 times better. Why?
Because hosting puts the runtime back in.
What users buy isn't that text—it's "that text being correctly executed." Execution requires machines, environments, accounts, and someone guaranteeing it runs to completion. None of that is a file; it's a service.
And services can be charged per use, per month, or per outcome.
That's the turning point of the whole story: the money isn't in the file; the money is in the runtime.
It's not a chatbot. A chatbot is "you ask, it answers"; a bot is "you state a task, it goes and does it, then comes back to you when it's done."
A personal agent typically looks like this: it has a name and a conversation thread that belongs only to you; it runs on a computer of its own, with a browser, file system, and terminal; it's connected to your email, calendar, cloud drive, code repositories, and payments; it remembers preferences you mentioned once; it can run on schedules and watch for events; and it stops to ask you before critical actions.
What it's actually used for now is mostly unglamorous: consolidating orders scattered across three platforms into one spreadsheet; watching a flight and alerting you when the price drops; turning meeting notes into a draft weekly report; haggling with customer service, unsubscribing, and processing refunds; scanning the inbox daily and drafting replies to what needs answering.
The common thread: it requires logging in, continuous operations, repeated execution, and someone to take responsibility when things go wrong.
This is exactly where skills can't reach. A skill teaches the model "how a weekly report should be written"; a bot is responsible for "actually writing the weekly report every Friday and sending it to you."
Representative products: Grok Bot is a team of named teammates sharing one cloud computer; Muse gives each person an isolated virtual machine usable right inside WhatsApp; dots is a resident agent inside ChatGPT that hands coding work off to Codex. The local camp is Hermes's Bot Mode and OpenBot, running on your own machine.
Back to the money question.
This wave of personal agents is, at its core, rebuilding the runtime that skill lacks — and selling it to you.
Breaking it down:
Skill has no execution entity. Muse gives each person a cloud VM, Grok Bot gives one account a cloud computer. The agent opens browsers, fills forms, runs commands — and keeps going after your laptop lid closes.
Skill has no account or credential relationships. Muse's credentials go into a vault, the model never sees real passwords, and payments go through one-time virtual cards; Grok Bot simply shares login state, trading security for convenience.
Skill has no run history. Each Hermes Bot has a permanent conversation; dots keeps making progress after you leave, and comes back to find you when it gets stuck.
Skill has no scheduling or triggers. Grok Bot saves proven workflows as routines; Hermes's scheduled tasks go straight into
hermes cron list
。
Skill has no accountable party. This is the most critical part. Muse pops up an approval before sending emails or spending money, with an outer layer called Sentinel governing network requests; Link backs agent purchases with compensation. These aren't features — they're accountability infrastructure.
So what a bot sells isn't a smarter skill. It sells: continuous operation, plus someone you can find when things go wrong.
One thing needs to be made clear here: this does not mean it's more reliable.
Bots make mistakes. And because they run continuously and progress across sessions, errors accumulate — when OpenAI beta-tested the resident version of Codex, its risk assessment explicitly noted this: drift carries down the entire task chain, and by the time a human comes back to look, the chain has already run far ahead.
Its advantage over skill isn't "making no mistakes," but that mistakes are traceable, interruptible, and compensable: approvals before actions, audit logs during execution, terms and compensation afterward, and a provider that is a corporate entity that can be complained about and regulated.
Accountability isn't a byproduct of capability; it's the price of charging money.
Users are willing to pay for it not because it's smarter than markdown, but because it can't run away.
Skill sells knowledge, and knowledge is depreciating in the model era; bots sell the fact that someone can be found — and finding someone has never been cheap.
At this point it's easy to slide into "bots are the answer." Hold on — there are three holes in that.
First, bots can be taken away too.
Hermes supports packaging and distributing an entire agent profile; OpenBot specifically wrote "How to Import Grok Bot Teammates"; nanoMuse open-sourced its mobile client, desktop client, and relay server entirely.
So "bots can't be copied" is false. The accurate statement is: the bot's package can be taken away, but it can't connect to your email authorization, payment channels, run history, and trust relationships. Taking away an empty shell is like taking away a new employee with no accounts.
The moat hasn't disappeared; it has just moved from "content" to "relationships and state."
Second, the cost structure is heavier, and gross margin isn't necessarily higher.
One VM per user, always-on compute, customer service, compliance, compensation — all hard costs. Muse relies on free tiers plus transaction commissions, Grok Bot is bundled into a subscription, dots is bundled into ChatGPT — they're all amortizing costs through existing distribution channels.
An independent company building a bot from scratch has to shoulder the compute bill alone. That's why almost all of this wave of products comes from big companies.
Greater user value does not mean greater business value.
Third, winner-takes-all still applies.
The bot market will most likely also see the top few take the majority. It's not a universal solution; it just shifts the competitive dimension from "whose markdown is written better" to "whose runtime, account system, and trust mechanisms are stronger."
Harder, but also more valuable.
Finally, this judgment has boundaries.
If you're building a private skill inside an enterprise, not planning to sell it, only using it for efficiency, then it never needed a checkout counter in the first place, and the argument above doesn't apply.
If your clients are buying your judgment rather than your files, then consulting delivery is still a good business — just don't mistake it for a product.
If you're a skill author:
Stop counting on making money by selling files. Three more realistic paths: treat it as a portfolio and lead-generation asset, charging for consulting, implementation, and customization; or turn it into a hosted service, selling runs rather than files; or accept that it's open source, trading it for reputation and opportunities.
If you're a platform:
Build the checkout counter next to the runtime, not next to the download page. Per-run, monthly, or revenue-share-on-results are all healthier than one-time buyouts. At the same time, be prepared to absorb "accountability" — approvals, audits, compensation. That's the part others can't copy.
If you're an enterprise decision-maker:
Don't ask "how much is this skill worth"; ask "if this goes wrong, who's responsible, how is it measured, can it be repurchased." If two of those three questions can't be answered, it's not a product — it's documentation.
What this wave of personal agents is really selling isn't intelligence.
Intelligence has already become cheap to the point of nearly free. What they're selling is a continuously existing relationship: it remembers you, it keeps working while you're away, it asks you before doing something wrong, and someone compensates you when it screws up.
None of these can a markdown file provide.
So stop agonizing over how to prevent others from copying your skill. The question you should be thinking about is a different one:
When knowledge itself is no longer scarce, what do you hold that others can't copy?
The answer most likely isn't what you know, but what you're willing to be responsible for.
53AI, the top service provider for enterprise LLM adoption
Products: scenario implementation consulting + LLM application platform + industry solutions
Promise: free POC validation, cooperation only after results meet the bar. Zero-risk LLM adoption, delivered to 160+ mid-to-large enterprises
Loading...
Scan to consult
Scan to log in
By logging in, you agree to the "53AI Website Service Agreement"
Service Agreement
Welcome to the [53AI Official Website] (hereinafter "this website" or "we"). This "Membership Service Agreement" (hereinafter "this Agreement") is a legal agreement between you (hereinafter "Member" or "User") and [Shenzhen Bosi Xiechuang Network Technology Co., Ltd.] regarding registration, login, and use of this website's membership services.
Before you register or log in, please be sure to carefully read and fully understand the terms, especially those exempting or limiting liability, intellectual property terms, and dispute resolution terms. Such terms will be highlighted in bold to draw your attention. When you successfully log in to this website via WeChat official account authorization, SMS verification code, or other means, you are deemed to have fully understood and agreed to accept the entire contents of this Agreement.
I. Definitions
This website: refers to the website operated by [Shenzhen Bosi Xiechuang Network Technology Co., Ltd.] with the domain [53ai.com], along with related mobile pages.
Membership Services: refers to the knowledge base article access, content search, and other related value-added services provided by this website to registered members.
Knowledge Base Content: refers to the digital content resources published on this website, including but not limited to text, charts, data, research reports, industry analyses, etc.
II. Account Registration and Login
Login Methods: This website supports the following login methods, and you may choose according to your actual situation:
WeChat Official Account Authorized Login: You agree to authorize your WeChat OpenID information to this website for the purpose of creating or linking a member account.
Mobile Verification Code Login: You must provide a genuine and valid mobile phone number and complete identity verification and login/registration via SMS verification code.
Account Security: Your account is for your personal use only; gifting, lending, renting, transferring, or selling it is prohibited. Any losses resulting from account theft or password leakage due to your improper safekeeping shall be borne by you.
Real-Name Verification: As required by relevant laws and regulations, we may require you to complete real-name verification for certain features. If you refuse to provide it, you may be unable to use some or all of the services.
Protection of Minors: If you are under 18 years of age, please read this agreement accompanied by your legal guardian and use this service only with the guardian's consent.
III. Service Content and Rules
Knowledge Base Access Rights: After logging in, members have the right to browse and search relevant articles and content in this website's knowledge base online, within the scope of permissions corresponding to their membership level.
Service Changes: We have the right to adjust, change, or terminate part of the service content based on business development needs, and will provide advance notice via website announcements, official account messages, and other means.
Prohibited Conduct: When using the services, you must not engage in the following:
Using technical means to bulk-crawl, download, or re-store knowledge base content;
Using knowledge base content for commercial purposes or disseminating it to third parties without authorization;
Interfering with the normal operation of this website or infringing upon the legitimate rights and interests of other users;
Publishing illegal or non-compliant information or engaging in activities contrary to public order and good morals.
IV. Intellectual Property Statement
Ownership of Rights: The intellectual property rights of the layout design, software code, and other content in this website's knowledge base belong to [Company Full Name] or the original rights holders, and are protected by laws such as the Copyright Law of the People's Republic of China.
Limited License: This website grants members a non-exclusive, non-transferable, non-sublicensable general license, limited to online access to knowledge base content for personal study and research purposes only.
Infringement Liability: Without written permission, no organization or individual may copy, repost, excerpt, mirror, compile, or otherwise use the above content in any form. Once discovered, we reserve the right to pursue legal liability.
V. Personal Information Protection
We value the protection of your personal information. For how we collect, use, store, and protect your personal information, please read the Privacy Policy separately.
The information you provide through WeChat official account authorization or mobile phone verification will be processed strictly in accordance with the Personal Information Protection Law, and used only for necessary purposes such as identity verification, service provision, and security verification.
You may at any time exercise your rights to access, correct, and delete personal information and withdraw your authorization consent through website settings or by contacting customer service.
VI. Disclaimer
Content Accuracy: Knowledge base content is for reference only and does not constitute professional advice. We make no express or implied warranties regarding its completeness, accuracy, or timeliness, and you should make your own judgment and bear the risks of use.
Force Majeure: We shall not be liable for breach of contract for service interruptions or delays caused by force majeure such as natural disasters, changes in policies and regulations, network failures, or third-party platform interface anomalies (such as WeChat interface maintenance or carrier SMS channel failures).
Third-Party Links: This website may contain links to third-party websites. The content and services of such websites are not under our control, and you should assess the risks yourself.
VII. Liability for Breach of Contract
If you violate this agreement, we have the right to take measures such as warnings, feature restrictions, service suspension, or account cancellation depending on the severity, and reserve the right to demand compensation for losses.
If your breach of contract causes us to suffer administrative penalties, third-party claims, or reputational damage, you shall bear full liability for compensation (including but not limited to fines, damages, attorney fees, notarization fees, etc.).
VIII. Governing Law and Dispute Resolution
The formation, performance, and interpretation of this agreement shall be governed by the laws of mainland China.
Any dispute arising from or in connection with this agreement shall be resolved through friendly negotiation between the parties; if negotiation fails, either party may file a lawsuit with a people's court of competent jurisdiction in [Company Location].
IX. Miscellaneous
This agreement constitutes the entire agreement between the parties regarding this service and supersedes any prior oral or written agreements.
If any provision of this agreement is deemed invalid or unenforceable, it shall not affect the validity of the other provisions.
We have the final right to interpret this agreement and reserve the right to modify it at any time within the scope permitted by law. The modified agreement takes effect immediately upon publication, and continued use of the service shall be deemed acceptance of the revised content.