Learn to give dot a small project, track changing deadlines and dependencies, and review its work using a fictional workshop example.
Guide stack
Most readers should leave with one of three next steps: a role guide, a prompt library section, or a course that matches the same problem.
Reader FAQ
If you want faster execution, open the prompt library. If you want a bigger decision, open the role guides or the course catalog.
Yes. Start with the guide hub, then use the sample lesson path or the prompt library before committing to membership.
Choose the next step that matches your job to be done, not the most popular page.
Put it into practice
Go from an empty folder to a finished habit tracker with Claude Code: first project, plan before you build, review changes, add tests, then write the README, save it with git and get it ready to publish.
5 lessons. US$20 once for this course. Also included with school access.
Lifetime access is one payment with no renewal. Membership is US$10/month or US$80/year. AI tools and their credits may cost extra.
Read next
AI Agents
How to Train Your Own AI Agent: Prompts vs RAG vs Fine-Tuning (2026)
6 min read
AI Agents
MCP Explained for Non-Technical Professionals: Why It Matters
11 min read
AI Agents
How to Use AI Agents for Small Business Workflows in 2026
11 min read
Guides
Learn AI by Finishing One Useful Project: Build Your Small-Business Website
6 min read
Updated September 30, 2026
A useful first OpenAI dot workflow has a modest goal: keep one small project understandable as its details change. Start with a fictional project, define what a good update contains, and check the result against the information you supplied.
OpenAI describes dots as agents that can continue working between conversations, using their own cloud computer and connected apps. That makes project follow-through a relevant use case. It also makes clear boundaries important. OpenAI’s dots introduction
This tutorial is a practice exercise, not a report of a live product test. Its business, people and records are fictional.
As of September 30, dots are rolling out to Pro users outside the European Economic Area, Switzerland and the UK. Business Premium availability covers supported ChatGPT regions. Enterprise users, including Edu and Healthcare, need an administrator to enable the beta. Rollout can take several days.
Create your dot through the ChatGPT desktop app or desktop website. Mobile cannot currently create a dot; the mobile app can provide access after setup when available. Mobile web is unsupported. Local computer access is optional and initially disabled. OpenAI’s getting-started guide
For this exercise, paste the sample information into the conversation. You do not need to connect a real inbox or give access to customer records.
Imagine a stationery shop preparing a weekend workshop. The project is ready when its room, materials and invitation draft are checked.
Use this fictional project brief:
Notice the distinction between a draft and an approved result. If that distinction is missing from the input, an attractive summary can conceal unfinished work.
Copy this prompt after the brief:
Help me track this fictional workshop project. Use only the information I pasted. Create a project summary with each task, owner, due date, status, dependency and missing evidence. Preserve the dates exactly. Mark unknown information as unknown. Treat September 30, 2026 as the planning reference date for this exercise.
Identify the next three useful actions and explain each using the brief. Keep proposed changes separate from confirmed facts. Draft questions for me to review, but do not contact anyone, publish anything, create calendar invitations or change connected apps. Finish with a short list of decisions I need to make.
This provides a goal, evidence and action limits. It also creates a way to judge the response beyond whether it sounds organized.
A satisfactory response should preserve all three owners and deadlines. It should identify room confirmation as a dependency for the invitation, without inventing a room or claiming that someone has replied.
Look for these specific checks:
These are expected properties of a correct answer, not results observed in a tool trial.
If a claim has no supporting input, ask:
Which line in my brief supports that statement? If none does, correct the summary and mark the information unknown.
Next, paste this update:
Fictional update: Advance the planning reference date to October 1, 2026. Morgan confirmed Studio B on October 1. Alex needs one extra day for the materials checklist; the proposed new deadline is October 4, but I have not approved it. Update the tracker. Show exactly what changed, what still needs a decision and which downstream task can now progress.
The room can become confirmed. The invitation dependency can be marked resolved. Alex’s original deadline should remain distinguishable from the proposed replacement.
This is a useful test because changing a date silently is easy. A dependable project summary preserves the difference between a request and your decision.
For a real project, choose a source you are authorized to share, confirm that dot can access it, and specify a finite review period. Avoid a vague instruction to monitor everything.
A reusable request is:
Review [approved project source] each weekday at [time and timezone] through [end date]. Report changed deadlines, unresolved dependencies and decisions needed in this conversation. Include evidence links where available. If the source is inaccessible, say so. Do not contact project participants or edit the source.
Check the saved instructions, timing and destination before relying on the schedule. In the desktop profile, Activity lets you inspect tasks and results. Scheduled contains recurring work. Pausing dot does not necessarily stop delegated tasks or future scheduled runs; inspect those separately when ending the workflow. OpenAI’s controls documentation
Keep a three-question checklist beside every project update: What changed? What evidence supports it? What decision is still mine?
After a week, compare the summaries with the original records. Record missed changes and unnecessary alerts. Adjust the instructions before expanding to another project.
For more practice with everyday business workflows, explore ChatGPT for Small Business, including lessons on customer replies and documenting operations. It provides related foundational practice; this article does not claim that the course includes dedicated dot training.