Build AI role-play training simulations in plain language. Describe the persona, scenario, and rubric, your coding agent produces a rubric-graded, SCORM-ready practice block for any LMS.
An interactive learning activity where a learner has a real, back-and-forth conversation with an AI-played character to practice a specific workplace skill, graded against an instructor-defined rubric.
From a plain-English description to a rubric-graded, SCORM-ready simulation. No scripting language, no templates, no XML.
Tell your coding agent the role, persona, situation, and stakes in a paragraph. No templates, no XML, no scripting language to learn.
Name the observable behaviors a successful learner demonstrates, weighted by importance. The AI grades every turn against your rubric.
"Make Sarah more skeptical." "Add an objective on decision authority." Iterate by talking with your coding agent, not by editing JSON.
One command produces a SCORM 1.2 zip. Upload to Cornerstone, Moodle, Canvas, TalentLMS, or any SCORM LMS. Scores post back to the gradebook automatically.
The same toolchain powers sales role-play training, customer service de-escalation drills, manager and leadership rehearsal, and compliance scenarios. Pick the role; the rubric does the rest.
Give every rep an unlimited pool of skeptical, distracted, or hostile buyers to rehearse pricing, discovery, and objection handling against, scored on the same rubric your enablement team grades real calls on.
Try a live sales role-playPut agents in the conversation, not in front of a training video. An AI customer reacts to their tone, empathy, and policy adherence in real time, so de-escalation gets rehearsed, not just watched.
Try a live support role-playLet new managers rehearse performance feedback, hiring interviews, conflict mediation, and terminations in private, with rubric-graded feedback on empathy, clarity, and follow-through.
Try a live management role-playTurn the same regulations into realistic conversations, HIPAA disclosures, conflict-of-interest talks, suitability checks, so learners rehearse the words they'll need under pressure, not just check a box.
Rehearse difficult patient conversations, breaking bad news, informed consent, motivational interviewing, with rubric-graded feedback on empathy, plain-language explanation, and shared decision-making.
Drop new hires into role-specific first-week scenarios, status updates, peer intros, escalation paths, so they practice the conversations of the job before week one, not just read the playbook.
When the session ends, the AI grades the full conversation against your objectives. Each one comes back with a weighted score, evidence pulled from the transcript, and a specific tip the learner can act on next time, every score auditable against the words that produced it.
Update objectives or weights any time. Re-export the SCORM package and the next attempt grades against the new rubric.
Weighted criteria evaluated turn-by-turn. Each response gets a breakdown by objective, not a single opaque number, so learners know exactly which behavior to work on next.
Monitor improvement across sessions and cohorts. Surface regressions early and identify which objectives consistently trip up learners across the program.
Track whether the learner actually moved the conversation to resolution, not just whether they said the right words along the way.
Tweak the persona, swap objectives, raise the difficulty, or rewrite the rubric, just describe the change in your own words. Your coding agent rebuilds the simulation and re-exports the SCORM package without you ever touching a config file. No proprietary DSL, no template format, no lock-in.
When you're happy, export as a SCORM 1.2 package for your LMS, or use the same HTML bundle standalone.
discover-authority (weight 3). Re-grading rubric.Paste any prompt into your coding agent to get a complete AI role-play training simulation as a single self-contained HTML file. Adjust the persona, rubric, or difficulty in plain language.
A tech lead practices pushing back on a PM who's adding work mid-sprint, without derailing team morale or sounding obstructive.
Using /edu-role-play, make an 8-turn role-play: a tech lead pushes back on a PM adding scope mid-sprint. Score on clarity, alternatives, and morale.
A manager rehearses a performance conversation with a defensive report who has missed quarterly goals. Empathy and specificity both required.
Using /edu-role-play, make a 12-turn role-play: a manager runs a performance talk with a defensive report. Score on specifics, empathy, and next steps.
An account executive handles pricing pushback from a skeptical enterprise buyer mid-call, with a competing quote already on the table.
Using /edu-role-play, make a 6-turn role-play: an AE handles pricing pushback from a skeptical mid-market VP. Score on discovery, value framing, and next step.
A product manager aligns engineering and sales leads who have conflicting priorities for the next quarter, without picking a winner.
Using /edu-role-play, make a 10-turn role-play: a PM aligns engineering and sales leads with conflicting priorities. Score on tradeoffs, decision, and write-up.
A support agent de-escalates an angry customer while staying inside the refund and SLA policy, words, tone, and timeline all graded.
Using /edu-role-play, make a 7-turn role-play: a support agent de-escalates an angry customer within refund policy. Score on acknowledgement, policy, and remedy.
A director mediates between two teams with overlapping deadlines and conflicting resource asks, without making either side feel overruled.
Using /edu-role-play, make a 14-turn role-play: a director mediates between two teams with overlapping deadlines. Score on neutrality, constraints, and plan.
Every AI role-play training simulation exports as a standards-compliant SCORM 1.2 package. Upload the zip, assign it like any other course activity, and completion plus score flow back to the gradebook automatically.
cmi.core.score.raw and cmi.core.lesson_status
Authoring, grading, LMS compatibility, data privacy, model choice, answered directly.