Guide Library

Week 1 Companion

Start With The Human, Not The Tool

Week 1Session Companion

  • 20–30 minGuided walkthrough

This guide is a direct companion to the Week 1 livestream. Follow along while you watch the live session or use this as a reference if you get stuck or miss some of the session and need to catch up.

What You Need for This Guide

  • Your role, responsibilities and working preferences to describe in the blank Profile.
  • A plain-text editor and a work-file location where you can save and reopen your Profile.
  • Your approved AI account for the setup check in step 5. You can create and save the file before that access is ready.

Choose the tool you’re allowed to use

Your approved tool

The tool your company approved.

Pick your approved tool above to see the steps and screenshots for it.

Standard Data rules for AI at work

Data Rules

Data rules for AI at work

Check with your IT Team before you paste any work data into an AI tool. Avoid names, account numbers, customer data, or anything sensitive without explicit permission from IT. If you aren't sure, redact anything sensitive and replace it with generic information. AI makes mistakes, make sure to fact check your agents outputs even if they sound right.

Before you start

  • Check with your IT teamAsk what your company allows before you put anything from work into an AI tool. Your company policy is the one that counts, and it is the one this library cannot see.
  • Use the tool your company approvedChatGPT, Claude, Microsoft 365 Copilot Chat and Gemini Enterprise are all covered here. Use the one your company approved, not the one you like best.
  • Sign in with your work accountA work account keeps your chats inside the agreement your company signed. A personal account in the same product sits outside it.
  • No approved account yetFollow the steps with made-up example text, or on paper, until IT gives you one.

Never put this into an AI tool

  • Customer or personal detailsNo names, email addresses, phone numbers, home addresses or account numbers. That covers colleagues as well as customers.
  • Regulated recordsHealth, financial, HR and legal records stay in the systems built to hold them, unless IT has told you in writing that this tool may see them.
  • Passwords, keys and access codesNo credentials of any kind, yours or anyone else’s, and nothing that grants access to a system.
  • Anything under an NDA or not yet publicUnreleased plans, contracts, pricing and source code need clearance before they go near a chat window.
  • Anything your company policy prohibitsYour company’s rules win over anything in this library. If policy keeps something out of AI tools, it stays out, whatever a guide suggests.

Do this instead

  • Swap real details for placeholdersWrite “Customer A”, “[client name]”, “$X”. You get the same quality of answer, and you put the real details back yourself.
  • Describe the situation rather than pasting the documentMost of what an agent needs is the shape of the problem, not the file it arrived in.
  • Trim a file before you upload or paste itTake out the identifying columns and pages first, or work from a short extract.

Check the answer before you use it

  • Read it as if a colleague had drafted itAn agent states wrong things with complete confidence. Your name goes on the work, so the check is yours.
  • Say where the agent helpedFollow your company rule on disclosing AI assistance, especially in client-facing or formal work.

When you are not sure

  • Ask before you paste or upload, not afterA question to your IT team takes a minute. Cleaning up after a data incident takes a great deal longer.
  • When in doubt, leave it outIf you would not put it in an email to someone outside the company, keep it out of the chat.
Open the full page
Download Guide
Start guide
5 steps · At your own pace
About saved progress

No account needed. Checkmarks save on this browser and device when browser storage is available. If storage is blocked or full, new checkmarks last until you reload or close this page. Progress doesn’t sync to other browsers or devices. Clearing this site’s browser data removes it, and private browsing may not keep it after you close the window.

Step

Create your agent files folder

  1. Check your starting point

    Your Agent Profile is a short file about your role, your organization and how you like to work. You'll save it and check how accurately your tool reads it so you can reuse that context in future chats.

    Over the six weeks you build one agent, a piece at a time. This week it learns your work from this profile. Later weeks add how you communicate, the information it can use and the work it can take on for you.

    Open Data Rules above and review best practices for handling your company data. If your tool approval is unresolved, the message below helps you ask. If approval is still pending, complete and save your profile locally through Step 4. Return to Step 5 when your approved account is available.

    IT Tool Request
    Hi [name or team],

I would like approval to use [AI tool name, for example ChatGPT, Claude,
Microsoft 365 Copilot Chat, or Gemini for Google Workspace] for my work as [your role].

What I would use it for: [one or two lines, for example drafting and editing
emails and documents, summarizing notes, working through problems, or setting
up an AI agent that understands my role and gives answers that fit my work].

A few specifics to make this easy to review:

1. I am asking for a company-approved work account on an enterprise or business
 plan, not a personal account.
2. I will keep out passwords, credentials, protected customer information,
 private employee data, confidential financial details, and regulated
 information. If we have a data-handling policy for AI tools please point me to it.
3. If [AI tool name] is not something we can approve, I am happy to use
 whichever approved tool we already have instead.

Until I hear back, I will not put any work information into the tool and will
not use a personal account.

Thank you,
[Your name]
[Your role / team]

  2. Create My Agent Files

    In a storage location you normally use for work files, create a folder named My Agent Files. Open the folder. Keep the current master copies of your Agent Profile and the other reusable files you build during the bootcamp here.

    Treat this folder as your source of truth. Update the master file when you accept a useful change, then replace any older copy you supplied to a tool. A chat does not update your saved files automatically. When changing tools, attach the relevant files or paste their text into the new tool and check how it uses them.

  3. Download and save your blank profile

    Select Download file below. Open the downloaded Agent-Profile-blank.txt in a editor and save your master as Agent-Profile.txt inside My Agent Files.

    YOUR RESOURCE

    Agent-Profile

    Click to Download

    Download file

    Make a small edit, save, close and reopen the file to confirm the change stayed. Keep the file in plain-text (.txt) format.

Step

Fill in your Personal DNA

Work in your blank Agent-Profile.txt. A sentence or two per field is enough.

  • My role: your title and the work that lands with you.
  • My main responsibilities: what you own, review or coordinate.
  • My current goals and priorities: what matters now. Update this when priorities change.
  • My recurring work: the meetings, reports and follow-up that repeat.
  • Decisions I regularly make: what you decide, recommend or escalate.
  • People I support or work with: the groups you work for and what they need.
  • My working preferences: how an answer should help you act.

You can be this specific without naming anyone. Describe the kind of work, the groups involved and what matters to you, and leave out names and confidential details.

Step

Fill in your Business DNA

Fill the seven Business DNA fields like you are explaining them to someone new to your workspace.

  • About my organization: what it does.
  • Products, services or functions: what your organization or team provides.
  • Customers and audiences: who it serves.
  • Important business priorities: what matters to the organization now.
  • Important terminology: define terms the agent could misread.
  • Important context and constraints: approval boundaries, operating limits and authoritative sources.
  • What the agent should never assume: keep the prewritten rule and add useful limits on guessing.

Step

Set your agent’s rules and finish your profile

Your profile now describes you and your organization. Set how you want the agent to work with that information, then finish and save the file you will test in Step 5.

  1. Review and tailor the operating rules

    Find How I Want My AI Agent to Work With Me near the end of your profile. The seven operating rules are instructions for how the agent should use your information, handle uncertainty and leave important decisions with you.

    Read each rule and keep all seven as your starting point. Think about one recurring task you listed earlier. If it needs a more specific instruction, add or refine a rule to fit that task while keeping the existing safeguards. If the rules already fit your work, leave them as they are.

  2. Finish the entries in your profile

    Read through your Personal DNA and Business DNA. Replace any remaining bracketed instructions with your own entries and delete unused list slots. If a fact that matters is still unknown, write not yet known so the agent can see the gap. Keep the First Setup Check section for Step 5.

  3. Save the version you will test

    Review the whole file using Data Rules above and remove anything that should not be shared. Save Agent-Profile.txt in My Agent Files, then close and reopen it. Confirm that your own entries and operating rules are present. This is the completed profile you will attach or paste into a new chat in Step 5.

Step

Run the setup check and try one small task

  1. Attach your saved profile

    Start a new chat in your approved tool. Attach your saved Agent-Profile.txt. If attachment is unavailable, open the file, select all its text, copy it and send it as the first message.

    Choose Add files and more, then Add photos & files.

    Example: the fictional profile and setup check sent together.

    Choose Add files, connectors, and more, then Add files or photos.

    Example: the fictional profile and setup check sent together.

    Choose Add and manage sources, then Upload images and files.

    Example: the fictional profile and setup check ready to send.

    Choose Upload & tools, then Upload files.

  2. Run the setup check

    The file includes the setup check, so the agent may already respond with a summary and questions. If it has, use that answer. Otherwise, copy First Setup Check below and send it in the same chat. Compare the answer with your file.

    First Setup Check
    1. Summarize what you understand about:
   - My role
   - My responsibilities
   - The people I support
   - My goals and priorities
   - My recurring work
   - My organization
2. Identify anything that appears unclear, incomplete, or contradictory.
3. Do not fill in missing information with assumptions.
4. Ask me up to five questions that would most improve your ability to help me with my work.

  3. Check the answer and update your profile

    Check one responsibility, one priority and one working preference. If the summary leaves out one of these, ask the agent to summarize that part of your profile before checking it. Correct unsupported claims in the chat. If a useful clarification belongs in your profile, add it to the master file yourself. A chat correction does not update that file. Leave genuinely unknown facts marked not yet known.

  4. Test the saved profile in a fresh chat

    Save the file. Open a fresh chat, attach or paste the saved version, and run the same check again. Check that any correction survived. If the first answer was already accurate, confirm the fresh answer still matches. If it is wrong, inspect both the saved wording and the agent’s interpretation. A vague reply does not by itself mean your profile is vague.

  5. Try your profile on one small task

    Stay in the chat where you tested your saved profile. Choose a small task where something in your profile should affect the answer, such as a short update for your usual audience or a checklist in your preferred format. If your organization doesn't allow real work details in the tool, use a made-up or non-sensitive version of your task.

    If you need an idea, send Find a quick profile task below and choose a suggestion you would use. The agent will point out the profile detail you can look for in the result.

    Find a quick profile task
    Help me choose one small task that lets me see how my Agent Profile can shape a useful answer. Use the profile in this conversation. If it is missing, ask me to attach or paste it and wait.

Suggest two or three tasks I could try in a few messages using a few notes or a short general description from me. Choose work an agent can draft, explain, summarize or organize in this chat. Each option should produce a small output, such as a short message, a brief explanation or a short checklist. Keep any output limits in my profile, including a tighter word or item count. Keep optional or conditional preferences optional in your suggestions and final task description. For example, keep “an analogy when it helps” conditional. Describe people by role or placeholder when suggesting what I would supply. Avoid broad plans, research, external actions and tasks that require sensitive details.

For each numbered option, state the small output, what I would supply, and one confirmed profile detail that would shape the result. Explain that benefit plainly, such as adapting to my audience or using my preferred level of detail. Name how the profile would change the wording, emphasis, format or level of detail. Restating my role, responsibility or the task itself does not show that benefit. Present options as possibilities, not pending tasks or projects you assume I have.

Include “Something else / none of these.” Let me choose a number, combine options or answer in my own words. Wait for my choice. If I reject the options, use my answer to find a better fit.

If the profile is too vague to explain a real benefit, ask one short question about an audience or working preference. Offer two or three possible answers and “Something else / none of these.” Wait, then use my answer. Do not ask me to rewrite my profile or invent missing details.

If I already named a suitable small task with a confirmed profile detail that can shape it, skip the suggestions. If my task is too large or unsuitable, offer two or three smaller chat tasks and wait for my choice.

Once I choose, give one short sentence naming the confirmed profile detail I can look for in the answer. Then give a one-sentence description of my selected task in one copyable code block. Write that description from my point of view, referring to me as “I” or “my”, never by name. Use only details I confirmed. Do not add a project connection, name, date or commitment I did not give you. Stop without doing the task or writing its brief. I will frame and send the task request next.

    Your profile supplies the recurring background. Now give the agent the details of this task using four parts, called the Context Sandwich. If you used the finder, put its task description under Task.

    • Task: what you need.
    • Context: what it needs to know about this situation beyond your profile.
    • Constraints: the rules it should follow.
    • Output Check: what the result should contain and what still needs your review.

    Before sending, name one thing your profile should change in the answer, such as its level of detail or the way it addresses your audience.

    Send the request and check the answer for that feature. Also check for details the agent missed or invented. Add a clarification to your master profile only if it will help with future tasks. Keep details that belong to this task in the chat.

Start With The Human, Not The Tool

Confirm your result

Confirm the outcomes below to finish this guide.

Final checks

Your own Agent Profile is saved as a reusable master file in My Agent Files, including any corrections from the setup check.

In a fresh chat using that file, you checked the summary against the profile and can identify what still needs clarification.

You can point to where your saved profile shaped the answer to one small task, and you know which facts from that task, if any, belong in your master file.

Provide Feedback

Thanks. That went straight to the people who write these guides.