Start by collecting real incidents: terse code reviews, late handoffs, or emoji misunderstood across cultures. Model triggers, assumptions, and missing context. Then craft branches that test curiosity, summarizing, and acknowledgement skills, rewarding language that slows reactions, invites specifics, and protects shared goals from personal defensiveness.
Disagreements change character across tools. A playful comment on video can read harsh in tickets. Include Slack threads, issue trackers, and meeting recaps, showing how timing, read receipts, and overnight delays shape interpretations, and when escalation to a quick call restores alignment without amplifying tension.
Remote collaboration mixes norms around bluntness, silence, and hierarchy. Scenarios should respect that complexity, presenting characters with varied comfort speaking up. Offer respectful phrasing options, private check-ins, and shared decision rituals that let quieter voices influence outcomes without forcing performative vulnerability or penalizing culturally grounded caution.