Step inside the moment.
ChronoForge turns a lesson into a story your students live: they investigate, decide under pressure, then compare their path with what really happened.
October 16, 1962. The White House. Click to play.
Every chapter is one real decision, played in six beats.
1
Arrive
A dated, full-screen moment. Something has just happened, and people are already arguing about it.
The storyteller writes the scene live from the teacher's facts and remembers every earlier choice.
2
Investigate
Follow leads: a letter, a map, a witness. Each holds a true, checkable clue that goes into your journal.
The clue is the teacher's verified fact; the AI performs it as a voice or a document.
3
Ask
Put a question to anyone in the room. They answer in character, knowing only what they could know then.
A character agent answers in real time, and VETTING checks every reply before you see it.
4
Work it out
Do the subject with the evidence in front of you: compute the rate, weigh the source, read the line closely.
The engine marks your answer, gives a hint after a first try and shows the worked reasoning; no AI grades you.
5
Decide
Every option is defensible and every option costs something. Or write your own move and see if the world allows it.
A referee judges your own move against the time and place; the engine, not the AI, moves the stakes.
6
Compare
At the end, your timeline sits beside the record: what really happened, why, what if, and how your class chose.
The AI writes your what-ifs from your path; a coach answers your reflection with one question.
Teachers own the facts. AI performs them, live.
The turning points are fixed, so every learner meets the same decisions and the class can compare. Between them, everything you say and do is answered in real time, and nothing the AI writes can move the facts.
Fixed by the teacher, the same for everyone
- The role you play and what you want
- The turning points, in order, with dates
- Every question and its options
- What each option costs
- The clues and their sources
- What you work out, and its answer
- What really happened, and why
Performed by AI, different every time
- Every line of dialogue, in voice
- How people react to your earlier choices
- Answers to your own questions
- Whether your own move is possible then
- Your ending, epilogue and what-ifs
- A coach for your reflection
Any subject with a decision in it.
A literature class plays Friar Laurence and compares its choices with Shakespeare's text. A science class follows John Snow's evidence through Soho. The anchor changes with the subject: the historical record, the source text, the evidence, expert practice.
Your lesson, their story, your discussion.
- Build a world from your lesson in minutes
- Describe the moment and paste your sources. The architect drafts the role, turning points, evidence and anchors; a reviewer checks it against the design rubric; you edit every word.
- Assign it with a six-letter code
- Students join on any device with their Google account and the code. No class lists to upload.
- See where your class split from the record
- Every student meets the same decisions, so their choices can be compared. Read their reflections and start the discussion where the class disagreed with history.
What should you recommend to the President?
- Strike the sites54%
- Quarantine Cuba29%
- Talk in secret17%
Ask: 54% chose "Strike the sites", but the record went with "Quarantine Cuba". What did the real actors know or fear that we didn't?
Designed around how people learn, not around a quiz.
Be someone, somewhere
Learners take a plausible role close to the decisions, with a want of their own. Perspective-taking needs a point of view, not omniscience.
Evidence before decisions
Clues and work-it-out questions pay: some options unlock only after the learner has found the evidence, or done the reasoning, that makes them possible.
Every choice costs something
Stakes are the values in tension at that moment. The engine does the arithmetic, so the numbers mean the same thing for every learner.
Failure you can learn from
If a stake collapses, the timeline breaks early and says why. Learners rewind to any chapter and try another path.
Compare, never grade
The record is a mirror, not an answer key. Matching history is not winning; the debrief asks why the real people chose as they did. Work-it-out questions give feedback, not scores.
Reflection with a coach
Learners explain their path in writing. A coach replies with one question that pushes for evidence or another perspective.
An open harness for AI in the classroom.
ChronoForge runs on the VIABLE Developer Platform from VIABLE Lab. The design rules live in one plain document, the world-design skill, that any AI agent can follow. The AI performs; teachers own the facts.
- Four published agents. Storyteller, Character, Reflection Coach and World Architect run on VIABLE, so their instructions, versions and traces are managed in one place.
- Vetted conversations. When learners talk to a character or a coach, a VETTING reviewer checks each reply for age, accuracy and the rule that nobody decides for the learner.
- A quality gate for new worlds. AI-drafted worlds are reviewed against the design rubric before a teacher sees them, then validated rule by rule.
- Research-ready. Decisions, evidence followed and reflections can join each learner's VIABLE record and the research export, under pseudonymous IDs.