ASSIGNMENT: PROMPT. EXPLORE. SHOW WHAT YOU CAN DO.
Working version: 19 November 2026
Improve after feedback by: 26 November 2026

GOAL
Use prompting to solve real problems in everyday life, study or work. Discover different useful things you can do, and learn to verify AI output. Try, inspect, improve and explain. Use your business project as a playground, or explore another useful task.

EVERYONE MUST SUBMIT
Submit at least one concrete experiment, even if it failed or your project is unfinished. A name or project link alone is not an experiment submission. Start today and add more experiments as you learn.

TRY DIFFERENT CAPABILITIES
Research, writing, image creation, presentations, websites, data analysis, integrations, automation, testing and debugging. These are examples, not a required tool list. Paid tools are not required.

FOR EACH EXPERIMENT
1. What real problem did you want to solve, and why did it matter?
2. Which tool did you use? Include the actual prompt.
3. What did it produce? Include an output excerpt and optional public evidence link.
4. How did you verify it? Show reliable sources, a manual calculation, a practical test or a check against your requirements. Record the result and limitations.
5. What failed? What did you change in your prompt or approach?
6. What became easier, faster or possible? What did you learn, and what will you try next?

SUBMIT
Open https://connectea-course-hub-vicorico17s-projects.vercel.app/join
Sign in at the course hub with your enrolled account, return, and connect your student session.
Paste your experiment log and choose Submit & get my agent rating.
Keep your own copy. Do not include secrets or other people's personal data.
Your name, submission status, score and feedback appear on the dashboard. Your full log is stored for the assignment and sent to the rating provider.

AGENT SCORE /100
25: real problems and useful exploration (15 for a clear need and useful outcome; 10 for breadth).
25: clear prompts and improved attempts.
35: verification and concrete evidence.
15: honest reflection and next steps.
One strong experiment can earn all 15 usefulness points; five distinct evidenced capabilities can earn all 10 breadth points. An unsupported claim that something works earns at most 5/35 verification points. Repeating the same task does not increase breadth. Failed experiments can demonstrate learning.
The agent reads your log; it does not open links or independently test your work. Include output excerpts and test observations. This is formative AI practice feedback, not an official university grade. It is not converted into the 35% Experimentation mark. Final assessment requires the individual evidence portfolio uploaded to the Year 4 assignment on the university’s Connect platform.

IMPROVE
Use the agent's next challenge to explore something new. Add evidence and resubmit (at least one minute between submissions). The latest submission determines your leaderboard score. If the agent is unavailable, your practice submission is still recorded and you can retry later.

DASHBOARD
https://connectea-course-hub-vicorico17s-projects.vercel.app/board
Every active student is listed, including Not submitted, Submitted / rating pending and Rated. Equal scores share a rank.
Optional project quests and repository signals are separate from the agent leaderboard.