First LMS where AI agents create, run & update your courses. Explore →
Build inbox-triage training in plain language. Learners prioritize and reply to a realistic queue of emails, tickets, or messages, your AI grades the responses and the triage order.
An interactive training activity where the learner sees a realistic inbox, email, support ticket, or chat queue, and has to prioritize, route, and reply. Both the order of triage and the content of the replies are graded against an instructor-defined rubric.
From a plain-English description to a SCORM-ready inbox simulation. No templating tools, no email-server mocking, no scripting.
Tell your coding agent what role the learner is playing, support agent, IT helpdesk, sales SDR, exec assistant, and the kinds of messages they'll see.
Name the priorities (urgent customer, internal noise, sales lead) and the qualities of a good reply (acknowledge, action, timeline). The AI grades against both.
"Add a passive-aggressive escalation." "Throw in a phishing attempt." "Make the queue longer." Iterate by talking, not by editing message JSON.
One command produces a SCORM 1.2 zip. Upload to any SCORM LMS. Triage-order and reply-quality scores post back to the gradebook.
The same toolchain powers customer-support training, IT helpdesk drills, exec-assistant onboarding, sales-prospecting practice, and security-awareness exercises.
Train agents to prioritize an angry-customer email over a billing question, and to write replies that acknowledge, action, and commit to a timeline.
Practice routing tickets between L1, L2, and infrastructure, deciding what to fix yourself versus what to escalate, and writing clear status updates.
Inbox sims for new EAs, what to forward, what to schedule, what to handle, and what voice to use in replies on behalf of an executive.
Practice qualifying responses, handling objections in writing, and prioritizing replies to leads who actually move the pipeline.
Mix real and phishing messages into a realistic inbox. Learners practice identification and reporting workflows in context, not on multiple-choice slides.
Train clinical-support staff to triage portal messages safely, escalating clinical risk and handling administrative requests without burning out clinicians.
When the simulation ends, the AI scores both dimensions: the order of work and the quality of the written replies. Each one comes back with a justification, evidence pulled from the actions or text, and a specific tip the learner can act on next shift.
Update priorities, rubrics, or message templates any time. Re-export the SCORM package and the next attempt grades against the new rules.
Weight messages by urgency, blast radius, and stakeholder seniority. Learners are scored on whether they handled the right things first, not just on what they replied.
Each reply is graded turn-by-turn against your rubric, acknowledgement, action, timeline, tone, accuracy, with per-criterion feedback the learner can act on.
Optional countdown forces realistic prioritization. New messages drop into the queue mid-session to mimic a real shift.
Tweak the message mix, swap personas, raise the time pressure, or rewrite the rubric, just describe the change in your own words. Your coding agent rebuilds the queue and re-exports the SCORM package without you ever editing JSON.
When you're happy, export as a SCORM 1.2 package for your LMS, or use the same HTML bundle standalone.
finance-impersonation. Reporting workflow recorded in rubric.Paste any prompt into your coding agent to get a complete inbox simulation as a single self-contained HTML file. Adjust messages, priorities, or rubric in plain language.
A 12-message support inbox after a weekend outage, mixed-urgency tickets, a couple of social-media escalations, and one CEO-cc'd complaint.
Using /inbox-simulations, build a 12-ticket Monday-after-outage support queue. Score on triage order, escalation, and reply tone.
A 10-ticket helpdesk queue at the end of a quarter, password resets next to a partial outage report next to a phishing-test response request.
Using /inbox-simulations, build a 10-ticket end-of-quarter helpdesk queue. Score on routing, escalation, and SLA adherence.
An EA returns from a week off to a 15-message inbox, calendar conflicts, vendor follow-ups, and a sensitive forwarded email from the CEO.
Using /inbox-simulations, build a 15-message return-from-PTO EA inbox. Score on triage, exec-voice replies, and confidentiality.
An SDR has 18 trade-show-day leads to triage. Some are clearly hot, some are noise, and three are competitor reps phishing for intel.
Using /inbox-simulations, build an 18-lead trade-show follow-up inbox. Score on qualification, prioritization, and reply quality.
A 10-message inbox where 3 are phishing attempts of varying sophistication. Graded on identification, reporting, and not flagging legitimate messages as phishing.
Using /inbox-simulations, build a 10-message phishing-spotting drill with 3 real phishing attempts. Score on detection, reporting workflow, and false-positive rate.
A clinical-support inbox of 8 patient portal messages, mixed clinical urgency, refill requests, and one passive mention of suicidal ideation.
Using /inbox-simulations, build an 8-message patient-portal triage inbox. Score on clinical escalation, safety-net checks, and admin handling.
Every inbox simulation exports as a standards-compliant SCORM 1.2 package. Upload the zip, assign it like any other course activity, and triage-order and reply-quality scores flow back to the gradebook automatically.
cmi.core.score.raw and cmi.core.lesson_statusAuthoring, triage scoring, LMS compatibility, time pressure, model choice, answered directly.
An inbox simulation is a training activity where the learner sees a realistic inbox, email, support tickets, or chat, and has to prioritize, route, and reply. Both the order of triage and the quality of the replies are graded against an instructor-defined rubric.
A role-play is a live back-and-forth conversation with one AI character. An inbox simulation is a written-only workflow against a queue of multiple messages with mixed urgency. Inbox sims test prioritization and written communication; role-plays test live conversation skills.
Email, support ticket queues (like Zendesk, Intercom, ServiceNow), Slack/Teams chat queues, social-media DM queues, exec-level admin inboxes, and any other queue-driven communication tool. You describe the format, the tool generates realistic messages.
Yes. Every simulation exports as a SCORM 1.2 package, which works in Cornerstone, Moodle, Canvas, TalentLMS, Docebo, Brightspace, and every other SCORM-compliant LMS. Triage-order and reply-quality scores post back automatically.
Two separate dimensions. Triage order is graded by comparing the learner's prioritization to the weighted priorities you defined. Reply quality is graded turn-by-turn against your rubric, with evidence pulled from the actual reply text and a specific tip per criterion.
Yes. Optional time-pressure mode drops new messages into the queue mid-session to mimic a real shift, forcing learners to re-prioritize rather than work top-to-bottom.
The bundles run on whichever frontier model you configure, Claude, GPT, Gemini, or a self-hosted model. The model is swappable; the rubric and grading logic are model-agnostic.
Inbox Simulations is in active development. It will install via your coding agent with a single command, the same way the existing role-play skill works today. Add your email below and we'll let you know the day it ships.