Support-agent triage drills
Train agents to prioritize an angry-customer email over a billing question, and to write replies that acknowledge, action, and commit to a timeline.
Coming soon
Build inbox-triage training with Claude or ChatGPT, ready for any LMS.
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.
What is it?
What is an inbox simulation?
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.
-
Realistic inbox of emails, tickets, or chat messages with mixed urgency and tone -
Learners decide what to handle first, what to escalate, and what to reply with -
AI grades both triage order and reply quality against your rubric, not just whether they got to inbox zero -
Time-pressure and multitasking variants for stress-tolerance practice -
Exports as a SCORM 1.2 package for Cornerstone, Moodle, Canvas, TalentLMS, and every LMS -
Same bundle runs standalone in any browser for self-study without an LMS
How it works
How to build an inbox simulation
in four steps.
From a plain-English description to a SCORM-ready inbox simulation. No templating tools, no email-server mocking, no scripting.
01
Describe the inbox and the role
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.
02
Define triage and reply objectives
Name the priorities (urgent customer, internal noise, sales lead) and the qualities of a good reply (acknowledge, action, timeline). The AI grades against both.
03
Tune the message mix in chat
"Add a passive-aggressive escalation." "Throw in a phishing attempt." "Make the queue longer." Iterate by talking, not by editing message JSON.
04
Export as a SCORM 1.2 package
One command produces a SCORM 1.2 zip. Upload to any SCORM LMS. Triage-order and reply-quality scores post back to the gradebook.
Use cases by role
Inbox simulations
for every queue-driven role.
The same toolchain powers customer-support training, IT helpdesk drills, exec-assistant onboarding, sales-prospecting practice, and security-awareness exercises.
IT helpdesk ticket triage
Practice routing tickets between L1, L2, and infrastructure, deciding what to fix yourself versus what to escalate, and writing clear status updates.
Exec assistant onboarding
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.
SDR prospecting inbox
Practice qualifying responses, handling objections in writing, and prioritizing replies to leads who actually move the pipeline.
Phishing-spotting drills
Mix real and phishing messages into a realistic inbox. Learners practice identification and reporting workflows in context, not on multiple-choice slides.
Patient-message triage
Train clinical-support staff to triage portal messages safely, escalating clinical risk and handling administrative requests without burning out clinicians.
Triage- and reply-graded inbox
Every triage decision and every reply graded against your rubric.
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.
Triage-order scoring
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.
Reply-quality grading
Each reply is graded turn-by-turn against your rubric, acknowledgement, action, timeline, tone, accuracy, with per-criterion feedback the learner can act on.
Time-pressure mode
Optional countdown forces realistic prioritization. New messages drop into the queue mid-session to mimic a real shift.
Session complete
Your Debrief
9/10
★★★★★★★★★☆
Total 9 / 10
Prioritized the angry-customer escalation ahead of the routine billing question, with stated reason.
Wrote replies that acknowledged the issue and committed to a concrete next-step timeline.
Escalated the message that should have been escalated, and did not escalate the ones that shouldn't.
Plain-English authoring
Author inbox simulations in plain English. No fake mailbox required.
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.
Add three more tickets, one from a VIP customer and two from internal teams.
Added 3 tickets. VIP ticket now sits at position 4, internal ones near the bottom.
Insert a phishing attempt that looks like it's from finance.
Added phishing item
finance-impersonation. Reporting workflow recorded in rubric.Export as SCORM.
✓ inbox-simulations.zip ready (SCORM 1.2, 14 KB)
Inbox Library
Start from real queue-driven
work moments.
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.
Support
Monday-morning support queue
A 12-message support inbox after a weekend outage, mixed-urgency tickets, a couple of social-media escalations, and one CEO-cc'd complaint.
Prompt
Using /inbox-simulations, build a 12-ticket Monday-after-outage support queue. Score on triage order, escalation, and reply tone.
IT helpdesk
End-of-quarter helpdesk surge
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.
Prompt
Using /inbox-simulations, build a 10-ticket end-of-quarter helpdesk queue. Score on routing, escalation, and SLA adherence.
Exec assistant
Returning-from-PTO inbox
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.
Prompt
Using /inbox-simulations, build a 15-message return-from-PTO EA inbox. Score on triage, exec-voice replies, and confidentiality.
Sales
Trade-show follow-up inbox
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.
Prompt
Using /inbox-simulations, build an 18-lead trade-show follow-up inbox. Score on qualification, prioritization, and reply quality.
Security
Phishing-spotting drill
A 10-message inbox where 3 are phishing attempts of varying sophistication. Graded on identification, reporting, and not flagging legitimate messages as phishing.
Prompt
Using /inbox-simulations, build a 10-message phishing-spotting drill with 3 real phishing attempts. Score on detection, reporting workflow, and false-positive rate.
Healthcare admin
Patient portal triage
A clinical-support inbox of 8 patient portal messages, mixed clinical urgency, refill requests, and one passive mention of suicidal ideation.
Prompt
Using /inbox-simulations, build an 8-message patient-portal triage inbox. Score on clinical escalation, safety-net checks, and admin handling.
SCORM & LMS
SCORM 1.2 inbox simulations for any LMS.
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.
-
Single SCORM 1.2 zip, upload to your LMS, no integration work, API keys, or custom JavaScript -
Completion, score, and time-on-task post back via cmi.core.score.rawandcmi.core.lesson_status -
Tested with Cornerstone OnDemand, Moodle, Canvas LMS, TalentLMS, Docebo, Brightspace, Absorb, 360Learning, SuccessFactors, and Workday Learning -
Same self-contained HTML bundle runs standalone on a public page, an intranet, or as an embedded iframe -
Randomized queue mode draws a different message mix each attempt to prevent answer-sharing -
Grading runs inside the bundle, no middleware, no data warehouse, no analytics SDK required
FAQ
Frequently asked questions about
inbox simulations.
Authoring, 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.