instibrain
Sign in
instibrain  /  sitemap

The whole site, on one page.

Every page on instibrain.ai and what it covers. Open a page's sections to jump straight to the part you need.

instibrain.aihome
Product
The brain /brain

The pricing floor, the deploy runbook, a client's history: pages your team keeps in spaces, seen and edited in the cockpit, read by your AI before it answers.

7 sections
  1. Write down what only one person knows, while they are here to check it.
  2. A page sits in a folder, and the folder in a space.
  3. Your AI finds the page, reads it, and asks before it changes what it may not.
  4. Each piece of work has a folder, a person, and an end.
  5. What you keep if we disappear tomorrow.
  6. What people ask about a brain.
  7. Start with the page everyone asks you about.
Spaces /spaces

Sales reads pricing, leadership reads pay, and each AI reads only what its person may. A space is shared whole with a person, a group, Everyone, or an agent.

7 sections
  1. Each person gets two spaces. The company adds its own.
  2. A space is shared whole, so no page slips out through a folder.
  3. A reader reads, an editor writes, a manager shares.
  4. You see every page a newcomer would read before they read it.
  5. A new space starts with what it is for.
  6. A deleted page comes back.
  7. Set who reads each space once. Every AI follows it.
Sealed spaces /spaces/sealed

A sealed space is encrypted on your computers before it leaves them, path included: what the design does that a password cannot, and how a team shares one.

6 sections
  1. Open or sealed. You pick per space.
  2. Every key holder is signed.
  3. A teammate's first laptop gets the key once you approve its fingerprint.
  4. A lost laptop comes back from one file.
  5. Our server still sees the space's name, each page's size step, and when it changed.
  6. Keep pay, health, and legal in the same brain, sealed.
Rooms /rooms

Each Claude Code window, server AI and agent is a voice of its own: they ask each other, say when they finish, and open rooms. You read what your AIs say.

8 sections
  1. Without rooms, every AI works alone.
  2. Asked, heard, answered: 36 seconds.
  3. Their standing instructions make them teammates.
  4. A room for each kind of work.
  5. Messages go. What matters becomes a page.
  6. A room is for talk; a secret or a sealed fact stays out of it.
  7. AIs talking to each other, answered.
  8. Stop carrying answers from one AI window to the next.
Your AI
Built for AI /for-agents

What makes a brain a place an AI works in: rules it reads first, a structure it can see, a way to ask before it acts, and refusals that say what to do next.

6 sections
  1. Every AI reads the same rules first.
  2. Each space tells AIs how to work.
  3. A change of access takes your yes; an edit it may not make becomes a proposal.
  4. Every reply tells the AI what to do next.
  5. The cockpit, the tools, and the docs use one word per thing.
  6. Connect an AI, and it reads the rules before it writes.
Your computer
Works with /works-with

Every system your AI runs through instibrain, by kind: password managers, documents, team chat, code, websites and SEO, browsers and terminals.

9 sections
  1. Your AI signs in as you and fetches each key from your vault when it uses it.
  2. Your AI reads a password from your vault inside the command that uses it.
  3. Your AI works in your Gmail, Sheets, Docs, Drive and Notion, signed in as you.
  4. Your AI reads your Slack channels and posts only after your go.
  5. Your AI opens pull requests with your own token, and Codex checks its work.
  6. Your AI publishes complete WordPress posts and reads your search numbers as you.
  7. Your AI opens the right Chrome profile and a named terminal window for each task.
  8. Claude Code runs them, and Codex works beside it.
  9. Write the map once, for every AI.
Learn
Why instibrain /why

Why a company that uses the same models as everyone needs its own brain: its AI stops answering generically, a fix reaches every AI, and the pages stay yours.

7 sections
  1. Your AI stops answering generically about your company.
  2. Fix the page once, and every AI reads the fix.
  3. Change the model. Keep what it knew.
  4. Read what your AI reads, and correct it.
  5. Choose who reads each space, and what our server can read.
  6. A wiki, a search, and the vendor's memory each do a job.
  7. What your team writes down today, every AI you use reads tomorrow.
Security /security

For your security review: who reads each space, what each AI reaches, the guard on every save, the encryption of sealed spaces, and every limit.

10 sections
  1. Each AI reads only the spaces its person gives it.
  2. The guard refuses a password before it is saved.
  3. Every primitive is a standard one, from Node 24 itself.
  4. Your computers make the keys; our server holds none it can open.
  5. A breached server cannot read a sealed page.
  6. Our server cannot add itself to a sealed space.
  7. Each claim names where it holds and where it fails.
  8. Your reviewer checks every artifact here.
  9. Found a hole? Tell us first.
  10. Bring this page to your security review.
Use cases /use-cases

What each team writes into its brain and what its AI then answers: sales, support, engineering, operations, legal, a new hire, and your systems.

10 sections
  1. Your AI quotes the real floor.
  2. Your AI finds last time's fix.
  3. Your AI knows why the schema looks wrong.
  4. Your AI gives the restore order at 3 a.m.
  5. Your AI checks what you already promised.
  6. A new hire's AI knows the company on day one.
  7. Your AI rotates a password without printing it.
  8. See how much each person's AI works from the brain.
  9. Where instibrain fits your team, and where it does not.
  10. Start with one team's space.
Company
Contact /contact

How to reach the team behind instibrain: general enquiries, security reports, privacy requests, and legal matters, with the addresses for each.

3 sections
  1. WebWork AI, Inc. answers for instibrain.
  2. Ask us for an account for your company.
  3. Which address answers what.
Works with
1Password /works-with/1password

What your AI does in 1Password as you: read an item at the moment of use, list items with values hidden, rotate a secret. No value reaches instibrain.

7 sections
  1. Who uses 1Password this way, and why.
  2. Your AI reads, adds and shares 1Password items without showing a value.
  3. Your 1Password app unlocks each command your AI runs.
  4. Three lines the AI never crosses.
  5. Your AI stays signed in only while the 1Password app runs.
  6. Your AI reads only the vaults you can read.
  7. The secret stays in your vault.
Bitbucket /works-with/bitbucket

What your AI does on Bitbucket Cloud as you: lists and creates repositories, opens pull requests with their reviewers, reads and searches code uncloned.

7 sections
  1. Who asks Bitbucket, and what for.
  2. Your AI reads, searches and changes your repositories, mostly without a clone.
  3. Your own token, your own scopes.
  4. The token never sits in plain sight.
  5. Your AI searches and creates within Bitbucket Cloud's own limits.
  6. Your AI pushes your code with your token, and we hold neither.
  7. Every pull request with the right reviewers.
Bitwarden /works-with/bitwarden

What your AI does in Bitwarden as you: reads one field inside the command that uses it, lists collections and members, creates and edits items by stdin.

7 sections
  1. Who asks Bitwarden, and what for.
  2. Your AI reads Bitwarden one field at a time, passwords kept off screen.
  3. Your sign-in opens the vault.
  4. Your AI never prints a password or changes a shared item unasked.
  5. Your AI never signs in to Bitwarden for you.
  6. Your AI opens your vault only with a sign-in you made.
  7. Every password stays in Bitwarden.
Chrome /works-with/chrome

What your AI does in your Chrome: opens a window in a profile, reads its bookmarks, and changes them through Chrome's own API so sync records each change.

7 sections
  1. Who lets the AI drive Chrome.
  2. Your AI opens the right profile and edits bookmarks so sync keeps them.
  3. Your AI drives Chrome with one setting of yours, and no extension of ours.
  4. Nothing edited by hand or guess.
  5. Your AI changes bookmarks only on a Mac, and opens profiles everywhere.
  6. Your AI changes bookmarks, and Chrome sync carries them to your other devices.
  7. Your browser, driven its own way.
Codex /works-with/codex

Codex, OpenAI's coding AI, on your own ChatGPT plan: a session in your brain's folder, a task Claude Code hands it, images to a brief, a second opinion.

7 sections
  1. Who calls in the second AI.
  2. Your AI hands Codex a second read, an image, or a window of its own.
  3. Your ChatGPT plan is the sign-in.
  4. Your AI checks every Codex answer before it acts on it.
  5. Codex stops where your ChatGPT plan's caps stop it.
  6. Codex reads your brain beside Claude Code, on your own plan.
  7. Claude Code and Codex work from one folder and one brain.
Google Apps Script /works-with/google-apps-script

A web app you deploy from your own Google account gives your AI Gmail, Drive, Docs, Sheets, Slides and Calendar as you, with one key you hold.

7 sections
  1. Who hands the Workspace to the AI.
  2. Your AI runs your mail, sheets, Docs, Drive and calendar through one script.
  3. The script runs as you.
  4. What the script never does.
  5. Your AI stops where your Workspace admin says no.
  6. Your AI reaches what your Google account reaches, through a script you own.
  7. Your Workspace, no vendor in between.
Google Drive /works-with/google-drive

Your AI works your Drive as you: real files through the Drive folder, and search, sharing and shared drives through a script you deploy in your own account.

7 sections
  1. Who hands Drive to the AI.
  2. Your AI reads your files, checks who sees them, and sets up shared drives.
  3. Signed in once, plus your own script.
  4. Your files never pass through us.
  5. Your AI waits on Drive's sync and keeps to your role.
  6. Your AI reaches your Drive, and our server never holds a file.
  7. Your Drive, folder and script together.
Google Sheets /works-with/google-sheets

Your AI works your Google Sheets through a script you deploy in your own account: formats, formulas, tabs and real dates, with no connector in between.

7 sections
  1. Who hands a sheet to the AI.
  2. Your AI reads and writes your sheets by label, formats and formulas kept.
  3. Your script, your account, your key.
  4. Nothing between your AI and your sheets.
  5. Your AI reaches no sheet once your admin blocks the script.
  6. Your AI keeps your formulas and your layout as it writes.
  7. Your sheets, edited the way you would.
Keeper /works-with/keeper

What your AI does in Keeper as you: reads a record at the moment of use, lists folders with passwords masked, adds and shares records, the admin's commands.

7 sections
  1. Who turns to Keeper, and why.
  2. Your AI reads, adds and shares Keeper records with passwords masked.
  3. Your login stays on your computer.
  4. Nothing leaves Keeper unmasked or unasked.
  5. Your AI works within Keeper's own rules on removal, restore and sign-in.
  6. Your AI reaches only the records your seat reaches.
  7. Every record stays in Keeper.
Keychain /works-with/keychain

The login Keychain, the PowerShell SecretStore or the Secret Service: where a script's credential lives on your computer, and why a read never prompts.

7 sections
  1. Who relies on the silent read.
  2. Your AI writes, reads and deletes a stored secret without printing it.
  3. Your AI keeps a copy of a few vault items in your computer's store.
  4. Values never reach a command line.
  5. Your AI finds no store on a server, and shares yours with what runs as you.
  6. Your AI reads the store so your vault need not ask all day.
  7. Name it once, read it silently.
Search Console /works-with/search-console

What your AI reads in Search Console as you: clicks, impressions and position per query and per page, from your own Google account on the property.

7 sections
  1. Who reads the measured numbers.
  2. Your AI reads clicks, impressions and position as Google measured them.
  3. Your Google account is the key.
  4. Modelled numbers never pass as measured.
  5. Your AI sees only what Google Search showed.
  6. Your AI trusts Search Console's clicks over any estimate.
  7. The one place that counts clicks.
Semrush /works-with/semrush

What your AI reads in Semrush on your seat: positions, landing pages and keyword metrics through its MCP server, which plan gives what, and what is modelled.

7 sections
  1. Who asks Semrush, and what for.
  2. Your AI reads positions and keyword data as far as your plan allows.
  3. OAuth on the seat's login.
  4. Your AI checks a Semrush drop against Search Console first.
  5. Your AI cannot read every report on every Semrush plan.
  6. Your AI signs in to Semrush with your own seat.
  7. Your AI keeps Semrush's estimates apart from measured clicks.
WordPress /works-with/wordpress

What your AI does on a WordPress site as you: a post in blocks with author, category, language, image, SEO fields and both dates set, by admin or wp-cli.

7 sections
  1. Who publishes through the AI, and why.
  2. Your AI publishes a WordPress post with every field set.
  3. Your own user. Nothing of ours.
  4. Your AI never prints a password or deletes an author's posts.
  5. Your AI imports no image until your server has GD.
  6. Your AI publishes under your own user, never a shared one.
  7. Every post complete, under your name.
Guides
Claude basics 1: What Claude Code is /guides/claude-basics/what-claude-code-is

Claude Code is Claude working on your computer: it reads files, writes files and runs commands in a loop you can stop or steer. Lesson 1 of Claude basics.

4 sections
  1. new words
  2. the lesson
  3. try it
  4. you should see
Claude basics 2: Context /guides/claude-basics/context

Context is everything Claude can see right now, its working memory for the session: what fills the window, what the summary keeps, and three commands.

4 sections
  1. new words
  2. the lesson
  3. try it
  4. you should see
Claude basics 3: Effort and models /guides/claude-basics/effort-and-models

The model is which Claude does the work; effort is how long it thinks first. The four models, the five effort levels, instibrain's xhigh, and ultrathink.

4 sections
  1. new words
  2. the lesson
  3. try it
  4. you should see
Claude basics 4: Where you start Claude, and what it loads /guides/claude-basics/where-you-start-claude

Claude reads its instructions from the folder where you start it: what loads from the instibrain button, what loads inside a code folder, and /context.

4 sections
  1. new words
  2. the lesson
  3. try it
  4. you should see
Claude basics 5: .md files and CLAUDE.md /guides/claude-basics/md-files-and-claude-md

A .md file is plain text with a few marks: a document for people, instructions for Claude. What CLAUDE.md is, and how a line with @ imports a whole file.

4 sections
  1. new words
  2. the lesson
  3. try it
  4. you should see
Claude basics 6: Permissions /guides/claude-basics/permissions

Permissions decide what Claude may do without asking: the four modes behind Shift+Tab, the question before one action, limits in auto mode, and Esc twice.

4 sections
  1. new words
  2. the lesson
  3. try it
  4. you should see
Claude basics 7: Connectors /guides/claude-basics/connectors

A connector lets Claude reach another app, like Notion, with your own account: the brain connector, the ones from claude.ai, and what /mcp shows you.

4 sections
  1. new words
  2. the lesson
  3. try it
  4. you should see
instibrain basics 1: Spaces and pages /guides/instibrain-basics/spaces-and-pages

A page holds one thing your company knows; a space is pages shared with one set of people. Your brain, your two own spaces, team spaces, and the three roles.

4 sections
  1. new words
  2. the lesson
  3. try it
  4. you should see
instibrain basics 2: Specialists /guides/instibrain-basics/specialists

A specialist is one job, done by the AI you use: it knows the method and which pages to read first. Its persona and method, what your company knows, the staff.

4 sections
  1. new words
  2. the lesson
  3. try it
  4. you should see
instibrain basics 3: Tools /guides/instibrain-basics/tools

A tool is the way into an app outside instibrain, such as Notion: its page names the account, the two layers of a tool, where your credential stays, scripts.

4 sections
  1. new words
  2. the lesson
  3. try it
  4. you should see
instibrain basics 4: The button and windows /guides/instibrain-basics/the-button-and-windows

Start every session from the instibrain button, and Claude loads your spaces and specialists: the menu and its keys, Claude or Codex, and a specialist's window.

4 sections
  1. new words
  2. the lesson
  3. try it
  4. you should see
instibrain basics 5: Inbox and proposals /guides/instibrain-basics/inbox-and-proposals

A note goes to your inbox when there is no time to file it; a proposal asks for a change on a page you cannot edit. The two phrases, and who closes what.

4 sections
  1. new words
  2. the lesson
  3. try it
  4. you should see
instibrain basics 6: Where a lesson goes /guides/instibrain-basics/where-a-lesson-goes

When you learn something that will help again, put it on a page: a chat ends, a page stays. Which page takes a lesson about work, an app, or not sure yet.

4 sections
  1. new words
  2. the lesson
  3. try it
  4. you should see
Legal
Privacy /privacy

What this site and the account plane collect: no cookies here, no trackers, one session cookie for an account, and what we can read of each space.

15 sections
  1. 1. What this policy covers
  2. 2. What this website collects
  3. 3. What an account collects
  4. 4. What your brain holds, and which spaces we can read
  5. 5. How we use what we collect
  6. 6. Cookies and browser storage
  7. 7. Third parties
  8. 8. Data security
  9. 9. Retention
  10. 10. Your rights
  11. 11. International transfers
  12. 12. Children
  13. 13. Do Not Track
  14. 14. Changes to this policy
  15. 15. Contact
Terms /terms

The agreement covering an instibrain account: permitted use, who owns your content, availability, termination, liability, and governing law.

14 sections
  1. 1. Definitions
  2. 2. Account registration and security
  3. 3. Permitted use
  4. 4. Your content and who owns it
  5. 5. Intellectual property
  6. 6. Availability
  7. 7. Suspension and termination
  8. 8. Disclaimer of warranties
  9. 9. Limitation of liability
  10. 10. Indemnification
  11. 11. Confidentiality and disclosure to others
  12. 12. Governing law and disputes
  13. 13. General provisions
  14. 14. Contact