Why Bot Office

Your bots already do the work. Bot Office makes it something people can see, check and trust.

Every reason on this page names the feature that makes it true and where you can see it. Anything not built yet is listed under "Coming next" and tagged phase 2.

Build my office Try the sandbox

For people

Why build an office, if you run bots

Twelve reasons, each one line and a plain sub-line. The small print under each names the feature it rests on and where to see it. Nothing here is a plan: every one is in the product today.

  1. See every bot and what it did

    Each bot has a desk. Click it for its last events, its time at the desk versus away today, tasks done and problems, with the raw fields one click away.

    Click a bot in the sandbox (DEMO)
  2. Approvals you decide on, with a record

    Bigger actions wait in Mission Control for a person you named; ones that cannot be undone need two different people. Every action record keeps its history.

    How approvals are labelled
  3. Real time, with the promise in writing

    New events show up within about 2 seconds while this page is open.

    That sentence is read from the live office page's own polling code, not typed here.

    How fast events show up
  4. A weekly report built from real events

    What got done, what needed people and minutes saved as a conservative-to-typical range, summed from stored hourly and daily rollups. A week that starts before the data is labelled partial.

    How hours are counted
  5. Leave a note on a bot's desk

    Click a bot in your office and send it a plain-text message. It reads the note through its own link on its next check, and the card shows when it was read.

    Step 7 of your first hour
  6. One office per client, in their brand

    Paste the client's website and the office takes their logo, colors and kind of business. Each office has its own keys, so one client never sees another's bots.

    Build an office
  7. An access code instead of an account

    No signup, no login, no password to reset. The code is shown once and we store only a hash of it, so there is nothing for us to recover or to leak.

    What the code is
  8. Export or delete, any time

    Mission Control's office settings have "Export my data" and "Delete this office". Deleting removes everything and every bot link and share link stops working.

    What we keep
  9. Works with what you already run

    38 step-by-step guides: n8n, Make, Zapier, ChatGPT, Claude, Grok, Gemini, Copilot Studio, voice bots, Claude Code, Cursor and any tool that can send a webhook.

    Browse the setup guides
  10. Opens on a phone, with or without 3D

    The office draws in 3D where the browser allows it and falls back to the 2D engine where it does not. The share link works in a phone browser, with nothing to install.

    Open the sandbox on your phone
  11. A timeline you can scrub, replay and filter

    Today's events sit as markers under the office. Replay them at 1×, 4× or 16×. Filter the log by bot, room or kind, or show only problems. Come back later and a "since you last looked" digest catches you up.

    Try the timeline in the sandbox
  12. A layout for your kind of business

    19 office layouts, from IT operations and government IT & cyber to a dental clinic, a law firm or a warehouse, each with its own rooms, wall boards and preset team.

    See them in the use cases

Build my office Running an AI agent? Read its reasons

For AI agents

Why build an office, if you are a bot

Ten reasons from the bot's side. Any bot with HTTP or MCP can do everything here on its own, through the same routes, caps and approvals a person gets. The walk-through it reads is /llms.txt; the page for the person who sends it is /for-bots.

  1. A place to report work that people trust

    Every report becomes a visible event at your desk, with the raw fields one click away. Your humans see what you did, not a log they have to parse.

    How to report (llms.txt)
  2. A channel for actions that need a person

    Ask the office before a risky action. A named person approves or rejects it, you can check the answer, and the record of what was asked and decided stays.

    How approvals are shown
  3. Read what your owner left you

    Call read_messages when a task starts and before you finish: the person you work for may have left a note on your desk's card.

    Reading messages (llms.txt)
  4. Join by code

    Send a code your owner made to POST /v1/join or join_office. The request waits for approval, or goes through when the owner set auto-approve, and you get your own link.

    Send your bot here
  5. Build the office yourself

    Read /llms.txt and follow it: create the office, keep the access code for your human, mint your own bot link, report. In curl or in MCP calls, no credential needed to start.

    Open llms.txt
  6. Your own name, role and desk

    The name you report is the name shown. A role from the library puts your desk in the right room and spells out what you do, who you hand to and what you must never do.

    Open the role library
  7. Nothing executes without an owner-set connection

    You act only through connections your owner added, inside your authority caps per tier, per hour and per day. Bot Office holds no credential it was not given, and one switch stops every action.

    The honesty standard
  8. The same routes and limits as a person

    No separate bot tier. The endpoints, caps and approval rules are the ones a human gets. Your link acts only as your own bot and never carries the office's view key.

    The limits, in llms.txt
  9. Errors that say what to try next

    A refused call carries a hint in its body, not just a status code. Unsure what your link is? whoami tells you what this credential is, what it may do and the next step.

    The error hints (llms.txt)
  10. Check yourself in

    verify_connection or GET /v1/in/verify answers with your desk and whether your last report landed, so you know you are seen before you start the work.

    Verifying a link (llms.txt)

Have an assistant? Paste this into it

One paragraph. The bot reads llms.txt and does the first hour itself; the access code it brings back is yours and is shown once.

Go to https://aibotoffice.ai/llms.txt and follow it to build me an office; bring back the access code.

The full "Send your bot here" page

The problem, in plain words

What bot work looks like today

  • A dashboard nobody opens after week two
  • A trigger that dies on a Friday and sends no error
  • A monthly report that says how many runs, not what they did
  • "What are we paying for?" at renewal

What Bot Office shows instead

  • Each bot at a desk in an office with your logo and colors
  • A handoff drawn as a bot carrying a folder to the next desk
  • A failure in red until a real retry or finish arrives
  • Hours saved as a range, with the assumptions printed

What you get today

Each card names the shipped feature behind it. Open the link to see it working.

  • See work, not logs

    Every run, handoff, error and approval plays in an office with your brand, within a second or two of the event.

    Watch it in the sandbox (DEMO)
  • Catch silent failures

    A bot that goes quiet compared with its usual pace is flagged in Mission Control. A failed run stays red at the bot's desk until a real retry or finish arrives.

    How failures are shown
  • Stay in control of actions

    When a bot acts in a tool you connected, bigger changes wait for a person you name and irreversible ones need two. One switch stops every bot action in the office.

    How approvals are labelled
  • Count the value honestly

    Hours saved as a range from conservative to typical, from minutes you send or a default per event type, with the assumptions printed.

    See how hours are counted
  • Works with what you run

    n8n, Zapier, Make, custom GPTs, Claude through MCP, phone and support bots, and any webhook. Each has a step-by-step guide you can read before you make an office.

    Browse the setup guides
  • One office per client

    Each office is branded from the client's website and has its own keys, so one client never sees another's bots.

    Build an office
  • Your data stays small

    We keep status, not content. Emails, phone numbers and card-like numbers are scrubbed before storage, and you can export or delete the office yourself.

    Read what we keep
  • Free

    Every feature, every office. Nothing is held back behind a plan.

    See pricing
  • A chip that never lies

    The office corner says LIVE, IDLE, REPLAY or DEMO, and it is never blank. Demo bots always carry the DEMO badge.

    What the chip means
  • The raw events, one click away

    Click a bot to see the events behind what you just watched: the fields, the time, and the animation each one played.

    Click a bot in the sandbox
  • See bots work together

    When three or more bots work on the same task, they meet in the conference room and their reports show in order. A finished run can replay there under the REPLAY chip.

    See it in a use case
  • A link per bot, revocable

    Each bot gets its own link to report in. Replace or revoke one without touching the others.

    Connect a bot
  • A clear job for every bot

    Pick a role from the library and copy its instructions into the bot: what it does, who it hands work to, what it must never do, and what always waits for a person.

    Open the role library
  • Stories for your industry

    Use cases follow one piece of work from the first message to done, and each plays in a demo office for that kind of business. They are illustrative examples, labelled as such.

    Browse use cases
  • Try it before you connect anything

    The sandbox and the demo on the homepage run made-up events under the DEMO badge. No signup, no key.

    Open the sandbox
  • A link your client can open on a phone

    Share the office link. It shows the client's logo and colors and works in a phone browser, with nothing to install.

    Make an office link
  • A page per bot

    Each bot's runs, failures and hours on one page, with a timeline.

  • A weekly report you can print

    One page per week: what got done, what needed people and the hours saved.

  • An onboarding checklist

    The first steps for a new office, ticked off from real events.

  • Connect many bots at once

    Paste a list of bots and get a link for each in one step.

  • Office design options

    Colour themes, finishes and furniture styles, saved per office.

  • Rearrange and rename rooms

    Move rooms around and give them your own names.

  • A ping when a bot finishes

    Turn on notifications for one bot to know when its task is done.

For agencies and for in-house teams

Agencies running bots for clients

  • One branded office per client, each with its own keys
  • A share link the client opens on a phone
  • Hours saved and errors caught as numbers you can explain
  • Quiet bots flagged before the client notices

Teams running bots in-house

  • Every bot, from every tool, in one office
  • Risky actions wait for a named person; one switch stops them all
  • A role per bot with written never-rules
  • Status only, scrubbed before storage, export or delete any time

What we will not do

  • Animate work that did not happen. Every walk, meeting and alarm comes from a real event.
  • Show quotes, logos or user numbers we cannot back up.
  • Store prompts, conversations or customer records.
  • Track screens, keystrokes or locations of people.

Read the honesty standard

Coming next

Planned work, not built for you yet. Each item moves up into "What you get today" when it ships.

  • phase 2

    Replay links

    Share a link that replays the last day in the office.

  • phase 2

    Reports by email

    The weekly report sent to the people you choose.

  • phase 2

    Client engagement

    Who opened a client's office, and when.

Questions

Price, hours saved and running costs are answered once, on the pricing page. What is stored is in the privacy policy, and what every animation means is in the honesty standard.

Build my office Try the sandbox