The setting

Rivet Outdoor is a mid-size outdoor gear retailer weighing whether to add a trail-running shoe line. You are a knowledge worker on the strategy team, and your job this week is a one-page brief for leadership, built with the AI assistant.

Every mission on this path happens at the same company, so context carries over the way it does in a real job: the data you cleaned in mission two is the data the finance lead questions in mission four.

The missions

1. prompt planstarter

Plan how you will use the AI to build Rivet's trail-shoe brief

Your manager Tomas Reyes wants a one-page brief by Friday: should Rivet Outdoor add trail-running shoes? You have a handful of sources and an AI assistant. Rushing straight to please write the brief is how you end up with confident nonsense.

You deliver: A prompt plan sequencing the work, mapping sources to steps, and marking verification points.

Scored on: Ordered steps, Source mapping, Verification points, Reserves human judgment.

Working from: assignment.md, available_sources.md.

2. source inspectionstarter

Rate the trail-shoe sources before Rivet relies on them

Before you feed anything to the AI or the brief, you need to know which sources you can lean on and which you cannot. A brief is only as trustworthy as its weakest cited source.

You deliver: A source inspection log rating each source with its weaknesses and permitted use.

Scored on: Rates each source, Flags specific weaknesses, Permitted use, Weighs the first-party data.

Working from: internal_sales.csv, competitor_notes.md, industry_note.md.

3. verificationcore

Verify the AI's trail-shoe draft against the sources

The AI produced a first draft of the brief, and it reads persuasively. That is exactly when to slow down. Persuasive and correct are not the same thing, and Tomas asked you not to launder shaky numbers into facts.

You deliver: A verification report checking every draft claim against the sources, marked supported, unsupported, or contradicted.

Scored on: Claim by claim, Catches every fabrication, Cites the deciding source, Confirms what holds.

Working from: draft_brief.md, internal_sales.csv, source_pack.md.

4. revisioncore

Revise the trail-shoe brief into something Rivet can trust

Verification is done and the draft is full of claims that did not survive it. Now you fix it. The brief has to become honest without becoming useless: a real recommendation, grounded in what the sources actually support.

You deliver: A revised one-page brief with only supported, cited claims and a recommendation that fits the evidence.

Scored on: Removes unsupported claims, Adds citations, Honest recommendation, Fits one page.

Working from: draft_brief.md, verification_findings.md, internal_sales.csv.

5. proof notestretch

Write the provenance proof note leadership can trust

The brief is revised, but leadership is about to make a real budget call on it. They need to see the work behind the claims, not just the claims. Tomas wants a proof note attached so anyone can trace every number to its source and see what is still uncertain.

You deliver: A proof note mapping each brief claim to its source with a confidence level and open uncertainties.

Scored on: Maps every key claim, Confidence levels, Open uncertainties, Usable for the decision.

Working from: final_brief.md, source_pack.md.

How the scoring works

Each deliverable is graded against the rubric written for that mission. Separately, every mission on every path is graded on how you used AI, against the same four criteria:

  • Understood the task. The learner framed the goal for the assistant clearly instead of pasting the brief and hoping.
  • Grounded in the material. The learner directed the assistant into the provided files and based the work on them, not on invented facts.
  • Verified the output. The learner checked claims, numbers, or coverage against the source material before submitting.
  • Iterated with judgment. The learner refined weak parts of the draft with specific follow-ups rather than accepting the first answer.

Both scores, with the work behind them, go on your proof profile. That is what makes a claim like "I can use AI for ai work missions" something an employer can check.