The honest answers.

A complete picture of your business — what you do, how you work, your processes, your team, your tools, your customers. Instead of re-explaining yourself every chat, your AI arrives knowing all of it. One-paste connection to Claude or ChatGPT, and it is ready to actually help.

Connect a structured knowledge base to your AI using a Claude Project or ChatGPT memory. The problem with most setups is that context disappears when the conversation ends. This gives your AI a permanent, structured file it can reference — covering your processes, team, tools, and how the business works — so you stop re-explaining yourself every session.

When you close the tab, ChatGPT forgets everything. The next conversation starts blank, so the first few minutes go to re-explaining who you are and what you need. This is that context — structured once, available every time. Your AI stops being a generic assistant and starts being one that knows your business.

The free plan generates your AI Action Plan — a prioritized guide to where AI can help your business first, based on your questionnaire answers. The paid plan unlocks your full knowledge base: a complete, structured brain your AI reads every time, plus a growing toolkit of skills it can run on your files. One-time payment, no subscription.

Both, and they work together. The documents are the brain: a structured account of how your business runs that your AI reads before it helps you. The toolkit is a set of skills your AI runs on that brain to do real work, like drafting an SOP or scanning for opportunities. The documents make your AI know the business; the toolkit puts it to work on the business.

Hosting is the plumbing, not the product. A folder gives you empty files. You get a structured business brain generated from six answers, a set of skills your AI runs on it, and a one-paste connection so your AI actually knows the business. A Dropbox folder gives you none of that.

The same reason you would pay for a business plan instead of writing your own: it is done for you, structured, and wired into your AI with zero setup. You are not paying for the files. You are paying for the brain we generate, the toolkit that keeps working on it, and the one-paste connection that keeps your AI current as the business changes.

No. The six questions are the same for everyone, but the brain is generated from your answers, so a plumber with three vans and a software founder with a remote team end up with different documents, different processes, and a different action plan. There is no form to fill in. We read what you tell us about the business and write the files around it, marking anything we inferred so you can correct it.

About five minutes of answering questions. We build the rest. You get a complete knowledge base generated and ready to connect — not a template to fill out, not a doc to write yourself. Five minutes of your time, and your AI is set up for every job after.

That structure follows the guidance Anthropic's own engineers and founders publish for giving AI context. The CLAUDE.md orients your AI; the knowledge base underneath it is the depth. We did not invent the pattern, we just set it up for your business so you do not have to.

Nothing. You answer six questions about your business. We generate your files. You paste one link into Claude or ChatGPT. That is the whole setup. The technical parts live behind the scenes.

No. You keep your existing AI subscription and keep using it the same way. This sets up what your AI knows — the context it reads before it helps you. Think of it as the briefing file your AI gets at the start of every conversation instead of starting blank.

Claude, ChatGPT, and Gemini. Drop the files into a Claude Project, paste them into ChatGPT memory, or connect via MCP if you use Claude Code. You bring the AI; we set up what it needs to know.

The files are plain text and easy to update — either yourself or by asking your AI to revise them. Because the context lives in structured files rather than inside the AI itself, keeping it current is a quick edit rather than starting over. The paid toolkit includes skills designed to help your AI maintain and update the knowledge base over time.

The files are yours. It is a one-time payment and you can download everything in plain text at any time. They work in any AI, any folder, any tool you choose. You are not renting access — you are buying a setup that lives wherever you want it.

The real cost is the generation pass when you sign up, and that is what the one-time payment covers. After that your own AI does the ongoing work, not ours, so there is no running meter on our side to fund. Your files are plain text you download and keep, so even in the worst case where we are not around, what you bought still works in whatever AI you use.

Your files are plain text stored under your account, downloadable any time. We generate your knowledge base once and then step back — your own AI does the ongoing work, and we are not sitting in the middle of your conversations. Our Privacy page has the full detail.

A CLAUDE.md is the orientation file an AI reads first to learn who it is working for. It is a plain markdown file that holds the essentials about your business: what you do, the tools you run on, who owns what, and how the AI should behave. Claude Code automatically loads a CLAUDE.md from the project root, and the same file works as Claude Project instructions or a pasted system prompt in any other AI.

You can write one by hand in any text editor, or generate one here in about two minutes. Answer six short questions about your business and we produce a structured CLAUDE.md for you, free. You get a real file you can download, edit, and drop into Claude, ChatGPT, or Gemini.

For Claude Code, put CLAUDE.md in the root of your project so it loads automatically at the start of a session. You can also place one in a subfolder for context specific to that folder. Outside Claude Code, paste the same content into your Claude Project instructions, into ChatGPT under a project or custom instructions, or at the top of any chat.

Paste the CLAUDE.md content into the Project's custom instructions or knowledge, and every chat in that Project starts already oriented to your business. The same file works as Claude Code's auto-loaded context when it sits in the project root, or as a pasted system prompt in any other AI. You write the context once and reuse it wherever you work.

Keep it short and high-signal. A good CLAUDE.md covers a one-line company snapshot, the core systems you use and what each is for, who owns what on the team, a pointer to deeper docs for detail, and a short set of instructions telling the AI how to behave. The goal is orientation, not a full manual. Depth lives in a linked knowledge base.

Yes. CLAUDE.md is just markdown, so it is not locked to Claude. The name comes from the Claude Code convention, but the file is portable. Paste it into ChatGPT, Gemini, or any assistant as instructions or memory, and it gives that AI the same context. You bring the AI; the file is what it reads.

No. Claude Code reads CLAUDE.md automatically from a project root, which is convenient if you use it. But the file is plain markdown and works anywhere. If you use Claude Projects, ChatGPT, or Gemini instead, you paste the same content in as instructions and get the same benefit.

Yes. Generating your CLAUDE.md is free, with no card required. You also get a free AI Action Plan and the first knowledge-base docs. The full knowledge base, the prompt and skill library, and the one-paste live connection are a one-time paid upgrade. There is no subscription.

Yes. The file is plain text and yours to keep. Edit it in any editor, or ask your AI to revise it as the business changes. Because the context lives in a file rather than inside the model, keeping it current is a quick edit rather than starting over.

CLAUDE.md is the orientation layer; the knowledge base is the depth layer. CLAUDE.md is the short file the AI loads first to get its bearings, and it points to the knowledge base for detail on your company, team, tools, products, customers, and processes. Together they let the AI start oriented and dig deeper when a task needs it.

The structure is the same; the difference is the work. You could write one yourself, the same way you could write your own business plan. We do it for you: six answers in, a structured CLAUDE.md out, organized the way Anthropic recommends for AI context, in two minutes instead of an afternoon of staring at a blank file.

An AI action plan is a short, prioritized document that tells a business where AI pays off first. Instead of a vague list of trends, it names concrete use cases for your specific business, ranks them by quick wins first, and ties each one to the tools you already run. The point is to turn AI from something you should look into into a clear next step you can act on this week.

You can write one yourself by listing your recurring work, marking where AI could help, and ranking those by effort and payoff. Or you can generate one here in about two minutes. Answer a few short questions about your business and we produce a structured AI action plan for you, free, with the use cases prioritized and matched to the tools you already use.

Yes. The AI action plan is free, with no card required. It stands on its own: you can read it, decide what to act on, and never pay anything. The optional one-time upgrade adds the full knowledge base, a library of skills and prompts, and a one-paste connection so your own AI knows your business. That upgrade is a single payment, not a subscription.

It covers where AI fits your business, a prioritized list of use cases with the quick wins first, each one tied to a tool you already run, a rough sense of effort and payoff, and a clear first move to start with. It is written to be read in one sitting, around 1,500 words, so you finish it knowing exactly what to do next rather than with more to research.

No. The plan is written for a business owner, not an engineer. It avoids jargon, names tools you already recognize, and describes each use case in plain language with the practical payoff. If you can describe what your business does in a few sentences, you can get a useful plan out of it.

A generic article talks about AI in the abstract. This plan is built from your answers, so the use cases are yours, the tools named are the ones you actually run, and the order reflects what will pay off fastest for your situation. It is a plan you can act on, not a trend piece you have to translate to your business yourself.

About two minutes. You answer a few short questions about what your business does, who is on the team, and which tools you use, and one pass turns those answers into a structured plan. There is no signup required to start and no call to book.

No. The action plan is a standalone free deliverable about how to put AI to work in your business. The knowledge base is a separate, deeper set of documents that describe the business itself so your AI can do real work on it. The action plan tells you where to start. The knowledge base, part of the paid upgrade, is what your AI reads to actually know the business.

You decide. The plan stands on its own, so you can take the quick wins and run with them yourself. If you want your own AI to know your business and act on the plan, the one-time upgrade generates the full knowledge base, adds a library of skills and prompts, and gives you a single link to paste into Claude or ChatGPT so your AI is grounded in your business from then on.

The plan is matched to whatever tools you tell us you run, the common ones for small businesses like your booking system, your inbox, your accounting software, your CRM, and your calendar. It points out where AI can sit on top of the tools you already pay for rather than asking you to adopt something new.

AI memory for a business is a durable, owned record of how your business runs that any AI assistant reads at the start of every chat, instead of relying on the limited memory built into the chatbot. You write the context once, the AI reads it every conversation, and from then on it works from your reality rather than a blank slate.

No. Built-in chatbot memory is small, automatic, controlled by the provider, and not portable, so it cannot hold how a whole business runs. It is aimed at light personalization: a preference here, a recent detail there. You cannot fully see or edit it, it is locked to one assistant, and it can forget or drift when a chat ends or a feature changes.

A memory file is a structured markdown document about your business that you control and edit, and it is portable across ChatGPT, Claude, and Gemini. It is the opposite of built-in memory: you decide what goes in it, you can read every line, it lives outside the model so it does not drift, and the same file works in whichever assistant you use.

A company snapshot, the systems you run on, who owns what, and how the AI should help. It is for orientation, not a full manual, so keep it short and high-signal and let a linked knowledge base carry the depth. That memory file is the front door to a larger structure we call an AI business brain.

Yes. The same memory file works in Claude, and there are Claude-specific ways to load it, into a Claude Project or a Claude Code project root, so it reads at the start of every conversation. We cover the Claude details in full on our Claude memory page.

You generate a memory file from a short questionnaire in minutes, with zero setup. Answer six questions about your business and one pass turns them into a structured memory file, organized the way Anthropic recommends for giving AI context. It is free to start, and you get a real file to download, edit, and drop into ChatGPT, Claude, or Gemini.

Generating your memory file is free, with no card required. You also get a free AI Action Plan and the first knowledge-base docs. The full knowledge base, the prompt and skill library, and the one-paste connection are a one-time paid upgrade. There is no subscription.

Partly. Claude has a built-in memory feature that can carry over some preferences and recent context, and it varies by plan. What it does not do is hold a structured, durable record of how your business actually runs. For that you give Claude a memory file: a plain markdown document about your company that it reads every time, so it starts oriented instead of guessing.

Not reliably on its own. Each chat largely starts fresh, and built-in memory is limited and not something you fully control. The dependable way to make Claude remember your business is to put the context in a file it loads at the start of every conversation. Then it knows your team, tools, and how you work without you re-explaining.

Both have a memory feature, and both are aimed at light personalization rather than a full picture of your business. The portable approach works the same way in either tool: a markdown file with your business context that you paste into Claude, ChatGPT, or Gemini. You write the context once and it gives any AI the same lasting memory of your company.

Put your business context in a structured markdown file and have Claude read it. You can write one by hand, or generate one here in about two minutes. Answer six short questions and we produce a CLAUDE.md plus a knowledge base about your company, free to start. Drop it into a Claude Project or Claude Code and Claude remembers your business from then on.

Claude's built-in memory is automatic, limited, and mostly out of your hands. A memory file is the opposite: you decide exactly what goes in it, you can read and edit it, and it lives outside the model so it never quietly drifts. For a business, control and accuracy matter more than convenience, which is why a file you own beats memory you cannot see.

With a memory file, yes. Because the context lives in a document rather than inside a single conversation, every new chat or project that loads the file starts with the same knowledge of your business. Put it in a Claude Project so all chats in that project share it, or in a Claude Code project root so it loads automatically.

A memory file makes it persistent and portable. The built-in feature can reset, change, or behave differently across plans and updates. A markdown file you keep does not: it is the same context every session, it survives model changes, and it moves with you to any AI tool. You own it, so it lasts as long as you want it to.

With a file, completely. You can open it, see every fact it holds, and edit or remove anything. Nothing is remembered that you did not put there. That is the main reason businesses prefer a memory file over automatic memory: there are no hidden notes about your company, just a document you control.

Yes. The memory file is plain markdown, so it is not locked to Claude. Paste the same content into ChatGPT custom instructions or a Gemini prompt and that AI gets the same memory of your business. You bring the AI; the file is the memory it reads.

Generating your memory file is free, with no card required. You also get a free AI Action Plan and the first knowledge-base docs. The full knowledge base, the prompt and skill library, and the one-paste connection are a one-time paid upgrade. There is no subscription.

Open ChatGPT, click your name, then Customize ChatGPT, and fill in the two boxes: what you want ChatGPT to know about you, and how you want it to respond. For a business, the first box is where your context goes: what the company does, the tools you run on, who owns what, and how you want it to work. The catch is that the box is small and gets summarized, so the better move is to keep your context in one structured file and paste the relevant part in. We generate that file for you in two minutes, free.

ChatGPT forgets because it has no memory of your business between sessions unless you give it that context every time. The model is general; it knows nothing specific about your company until you tell it. Custom instructions and memory help a little, but they are short, get summarized, and reset across projects and accounts. The reliable fix is to keep your business context in a file you paste in, so every chat starts oriented instead of blank.

Not on their own. ChatGPT custom instructions are a small box meant for short preferences, so they cannot hold everything an AI needs to actually work for your business: your systems, your team, who owns what, your real processes. They also get summarized and do not carry across projects or other tools. For a business, a structured context file you paste in holds far more, stays consistent, and works everywhere. Custom instructions are the on-ramp; the file is the real thing.

Custom instructions are a short preferences box inside one ChatGPT account; a business context file is a portable markdown document you own. The box is capped in length, gets summarized by the model, and does not follow you to Claude, Gemini, or a teammate's login. The file holds your full operational context, stays exactly as you wrote it, and works in any AI. You write it once and reuse it everywhere instead of re-typing instructions into every tool.

Yes, if you keep them in a file rather than in one tool's settings. ChatGPT custom instructions live inside ChatGPT and do not transfer. A plain markdown context file is portable: paste it into ChatGPT custom instructions, into a Claude Project or Claude Code, or into Gemini, and each one gets the same business context. You bring the AI; the file is what it reads. That is exactly what we generate, so you are not maintaining separate instructions per tool.

Lead with the essentials your AI needs to do real work: a one-line description of what the business does and who buys, the core tools you run on and what each is for, who owns what on the team, and how you want the AI to behave. Keep it high-signal and skip filler. Because the custom-instructions box is small, the practical approach is to keep the full version in a structured file and paste in the part that fits. We generate that structured file from six short questions.

Each custom-instructions box holds about 1,500 characters, and there are two boxes: one for what ChatGPT should know about you and one for how it should respond. That is only a few short paragraphs each, and the model summarizes what you write rather than using it word for word. For a business that fills up fast, which is why keeping your full context in a file and pasting in the part that fits works better: you sidestep the cap and the wording stays intact.

A good set is short and high-signal. In the first box: one line on what the business does and who buys, the core tools you run on, who owns what on the team, and your current priorities. In the second box: tell it to assume that context, ask before assuming details it does not have, be concise and lead with the answer, and flag anything that needs your decision. Keep it tight, because the box is capped at about 1,500 characters and gets summarized. The fuller version belongs in a context file you paste in and reuse across tools.

Yes. For the first box: "I run [business], a [type] that [what you do] for [who buys]. We operate on [core tools]. Team: [who owns what]. We are focused on [current priorities]." For the second box: "Assume the business context above, ask before assuming details you do not have, be concise and lead with the answer, and flag anything that needs a decision from me." That fits the limit and works as a starting point. The file we generate is the same idea with real depth, and it is portable to Claude and Gemini.

Yes. Answer six short questions about your business and we generate your context file for free, with no card required. You also get a free AI Action Plan and the first knowledge-base docs. The full knowledge base, the prompt and skill library, and the one-paste live connection are a one-time paid upgrade. There is no subscription. The file is plain markdown you can paste into ChatGPT custom instructions, Claude, or Gemini.

No. You can write them by hand, the same way you could write your own business plan, but the point is that it is done for you. Six answers in, a structured business context file out, organized for AI to actually use, in two minutes instead of an afternoon staring at a blank box. The file is yours to keep and edit, so you can still adjust it by hand whenever the business changes.

Custom instructions apply to new chats in the account where you set them, but not everywhere you might expect. They can be overridden inside projects, they get summarized rather than used verbatim, and they do not carry to other accounts or other AI tools. A context file you paste in gives you control: you decide exactly what the AI sees at the start of a conversation, and it reads the same in ChatGPT, Claude, or Gemini.

The structure is the same; the difference is the work and the portability. Hand-typed custom instructions live in one tool and tend to be thin because the box is small. We generate a structured file from your answers, organized the way an AI can actually use it, then you paste it wherever you work. You get the depth a real business needs and one source of truth instead of scattered instructions in each tool.

An AI business brain is a structured set of markdown documents, a CLAUDE.md plus a knowledge base, that gives any AI assistant the operational context to do work for your specific business. You write the context once, the AI reads it at the start of every conversation, and from then on it works from your reality instead of generic guesses.

A general AI assistant was trained on the public internet, not on your company, so it has no memory of your services, your team, your pricing, or the tools you run on. Every chat starts from zero, which is why you re-explain yourself each time. A business brain hands the AI those specifics up front so it stops guessing.

Two layers of context and a toolkit that runs on them. The CLAUDE.md is the orientation layer your AI loads first: what the business is, the systems you run on, who owns what, and how it should help. The knowledge base is the depth layer behind it. The toolkit is the skills, agents, and prompts that turn that context into work.

Operational context is context an AI can act on; marketing context is copy that describes the company. A marketing blurb tells a reader what you sell. Operational context tells the AI how the business runs: which tool holds the bookings, who approves a refund, how a quote gets built. A business brain is written for the second job, so the AI can do the work rather than talk about it.

Custom instructions and built-in memory are small, light, and mostly out of your hands; a business brain is structured, deep, and yours to edit. A few lines or an automatic note can nudge tone or recall a preference, but they were not built to hold how a whole business runs. The brain is a file you own, deep enough for the AI to act on, and portable across tools.

Small businesses and solo owners who want AI to do real work without a setup project. If you run the business and you are the one using AI to draft, plan, and reply, the brain is what makes that AI useful instead of generic. You do not need a technical team or a long onboarding.

You generate one from a short questionnaire in minutes, with zero setup. Answer six questions about your business and one pass turns them into a structured brain: a CLAUDE.md, a knowledge base, and a toolkit, organized the way Anthropic recommends for giving AI context. It is free to start, and you get real files to download, edit, and drop into Claude, ChatGPT, or Gemini.

Yes. The brain is plain markdown, so it is not locked to one tool. Paste the same context into ChatGPT custom instructions or a Gemini prompt and that AI gets the same operational picture of your business. You bring the AI; the brain is the context it reads.

Generating your brain is free, with no card required. You also get a free AI Action Plan and the first knowledge-base docs. The full knowledge base, the prompt and skill library, and the one-paste connection are a one-time paid upgrade. There is no subscription.

Custom instructions are fine for tone and small preferences. A structured folder of docs is what you need when you want AI to actually run your business. Custom instructions are a few lines in one global setting; a folder of docs is a CLAUDE.md plus a knowledge base, deep enough to hold how the business runs and portable across tools.

Custom instructions are a short block of standing guidance you set once so the AI applies it to every chat. They are a single, global setting, usually a few lines, best for steering tone, format, and simple recurring preferences. They are quick to set up, but they were not built to hold the structure of a whole business.

A folder of docs is a CLAUDE.md plus a knowledge base you own and edit: a structured set of markdown files that give any AI the operational context to do work for your business. The CLAUDE.md orients the AI first, the knowledge base carries the depth, and the whole thing is portable across ChatGPT, Claude, and Gemini.

When you use AI solo and mostly need it to match your voice, keep a consistent format, or remember one simple preference. If the job is tone and light personalization rather than running operations, custom instructions are the faster, lighter answer, and a folder of docs would be more than you need.

Because it holds how the business runs, so the AI can do work, not just sound right. Custom instructions can shape how the AI talks; a folder of docs gives it the systems, owners, pricing, and processes behind the talk. That is the difference between an assistant that sounds on-brand and one that can actually build the quote and follow your process.

You generate one from a short questionnaire in minutes, with zero setup. Answer six questions about your business and one pass turns them into a structured folder of docs, a CLAUDE.md and a knowledge base, organized the way Anthropic recommends for giving AI context. It is free to start, and you get real files to download, edit, and drop into Claude, ChatGPT, or Gemini.