> ## Documentation Index
> Fetch the complete documentation index at: https://docs.wejam.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Creating a Mission in Jam Studio

> Author a roleplay mission with a scenario, tasks, scorecard, and AI persona.

## Overview

A mission is one roleplay with one learning goal, made of three parts: the **persona** the learner talks to, the **scenario** that sets the scene, and the **scorecard** that defines what a good conversation accomplishes. You edit all three in the mission editor - "Magic mode" - reached from any draft track: open the track in [Jam Studio's](/external/coach/user-documentation/creating-publishing-tracks) **Drafts** tab (visible with the **Editor** or **Admin** [role](/external/coach/user-documentation/assigning-roles)) and click the pencil on a mission. New missions start from an existing one via **Create similar mission** in the mission's row menu.

## Working in the editor

The editor shows the mission's title, learning outcomes, and difficulty, with the three building blocks below and the **Maker Assistant** alongside: describe any change in plain language ("make the customer more skeptical", "rewrite this for the German market") and it applies the edits for you - attach context documents to ground them. Every change is preview-only until you save, and **Test mission** lets you play the roleplay yourself before learners do.

## Setting up the persona

The persona section defines who the learner speaks with: name, profile, workplace, and workplace description, plus gender and voice - preview any voice with the play button. The **Detailed persona profile for the AI** field holds background only the AI sees, which is what makes the character respond consistently under pressure. Field icons mark what's **learner-facing** (shown in the briefing) versus **AI-facing** (invisible to learners); toggle **Learner view** to check exactly what they'll see.

## Writing the scenario

The scenario tab holds the situation description the learner reads, background information for the AI, and the **conversation steps** - the outline the persona follows through the call, including how to react to what the learner does. Write the description as the learner's mission brief: who they are, who they're calling, and what they need to achieve.

## Defining the scorecard

The scorecard lists the tasks the learner must accomplish - each becomes a verdict on the [feedback page](/external/coach/user-documentation/feedback) after every round. Keep each item one observable behaviour ("Confidently introduce yourself and ask to be transferred"), expand **Show details** to refine the evaluation criteria, and use **Add scorecard item** for new tasks. Fewer, sharper items beat a long checklist: each one should be coachable on its own.

## What happens next

* Save your changes, then **Test mission** to play a round yourself - the fastest way to catch a persona that breaks character or a scorecard that can't be satisfied.
* The mission goes live for learners when its track is [published](/external/coach/user-documentation/creating-publishing-tracks).
* Learner attempts are scored against your scorecard items - managers [coach from those verdicts](/external/coach/user-documentation/coaching-with-scorecards), so wording matters.

## Related articles

* [Creating and Publishing a Track](/external/coach/user-documentation/creating-publishing-tracks) - the track your mission lives in
* [Understanding Your Feedback](/external/coach/user-documentation/feedback) - how your scorecard surfaces to learners
* [Coaching with Scorecard Results](/external/coach/user-documentation/coaching-with-scorecards) - how managers use what you define
* [Starting and Completing a Roleplay](/external/coach/user-documentation/roleplay) - the mission from the learner's seat

***

*Last updated: June 2026*