Quest Ladder
TeachFORGE.ai
Quest Ladder knowledge base
The Quest Ladder
Design a question once, let it generate endless versions of itself, arrange those into a path, and decide how much of that path each class gets.
You set the direction. The delivery sets the dose.
Four things, in order
Quest Ladder has four building blocks. Almost everything else in the menu is a way of making, checking or sharing one of them, so it is worth being sure of the order before you start.
1
A question type
The design. What is being asked, which values change between versions, how it is marked, and what a learner sees after they answer.
2
A question
One generated instance of that type, with its changing values filled in. One type can produce hundreds.
3
A ladder
An ordered path through several question types, with the rules for moving between them.
4
A delivery
A ladder handed to learners in a particular way: an exam, an assignment, a sprint, a live class round.
The ladder owns the direction — which types, in what order, with what branching. The delivery owns the dose — how many of them each learner actually sees. That split is why the same ladder can be a gentle homework and a timed exam without being rebuilt.
The guides built into the product
Six longer guides live inside the app itself, linked from the screens they explain. They go deeper than these articles on the parts where the engine makes a decision for you.
- Quest Ladder engine guide — how a path is chosen, what a gate mode does, and what happens when a learner is not ready.
- Exit mechanism routing guide — which learning block runs after each combination of outcome and confidence.
- Learning journey guide — skill status, this visit, and when a review falls due.
- Rigor levels guide — what counts as learned, and how strict you want to be.
- Manual build guide — changing values, rules, expressions and smart add-ons.
- Visuals without uploading images — diagrams, charts and maths drawn from text.
Everything in this section
Question types
- Question types: the list pageFind a type among hundreds, see at a glance which ones are finished, and know what deleting one takes with it.
- Manual build, Assisted build and AI buildThree ways to write a question type: by hand, with AI on every step, or in one pass.
- Inside the question type editorStages, steps, the seven section chips, and the dock that tells you why saving is blocked.
- Question details: basics and classificationThe format, the description the AI reads, and the classification that decides where a type turns up later.
- Presentation: how a question is styledPlatform default, a named pack, or the plain baseline — and why this changes typography rather than layout.
- Dependencies: finish first, unlocks, groups and placed underSay what should come before this type and what it leads to, so a ladder can route learners sensibly.
- The Quest Ladder dashboardYour counts at a glance and the four quick actions that start most sessions.
Question formats
- The twenty question formats at a glanceWhat each format asks a learner to do, how it is marked, and which article covers it.
- Choice formats: single answer, select all, true or falseThree formats where the learner recognises rather than produces, and the penalty rule that catches people out.
- Blanks and cloze: fill in the blank, multiple blanks, cloze passageMaking the learner produce the answer rather than recognise it, and how each blank is marked.
- Numbers, equations and gridsNumeric input, equation input and matrix grid — tolerance, units and equivalent forms.
- Short answer, long answer and proofThe three formats that ask a learner to write, and which of them reach your marking queue.
- Matching, ordering and drag and dropThree formats about relationships and sequence, with decoys, positions and zones.
- Hotspot on image and diagram labellingTwo formats that ask a learner to point at or name parts of a picture.
- Multi-step questionsOne problem worked through in stages, with marks at each stage.
- Code execution questionsCode a learner writes that must actually run, checked against test cases.
- Graph sketch questionsA learner draws a curve or plots points, and you mark what they drew.
- Extra settings: the knobs each format addsEvery per-format setting, its range and its default, in one place.
Settings and marking
- Shared settings: scoring, timing and hint timingThe seven tabs of settings every generated question inherits, and how a final mark is actually composed.
- Answer policy: how an answer is acceptedMatch modes, evaluation modes, tolerance, ranges and the alternatives list.
- The correct answer step: slots, marking mode and penaltiesGraded slots, all-or-nothing against weighted, splitting marks and the acceptance rules.
- The mastery gate: learned today, still known laterHow much evidence counts as having learned a topic, and what a strong learner should get instead.
- The repetition schedule: when to check they still rememberIntervals in days, how an exam date bends them, and what happens when a memory check fails.
- Rigour levels and Quest Ladder settingsWhat Standard means for your own work, and how to change it once for everything.
Dynamic content
- Dynamic variables: one type, endless questionsThe eleven parameter types, the rules that keep a generated question sensible, and the tokens that place a value.
- Expressions and computed factsComputing a value from your variables, and why the answer should be derived rather than typed.
- Data tables: look a value up instead of computing itPaste a table, key it, and let a question read a real value out of it.
- Shortcut placeholdersYour own chips on the Insert tray's quick row, for the values you place over and over.
- Smart add-ons: logic and repeats inside a questionNamed little routines that build a piece of content from your variables, and the ten step kinds they are made of.
- The Insert trayThe chip bar above every content field, and what each of its six dropdowns holds.
Content
- Content blocks: text, maths, media, diagrams, interactiveThe twenty-one block types a question is built from, grouped, with what each is for.
- Visuals without uploading imagesDrawing diagrams, graphs and mathematics from text, and why that beats a screenshot.
- The question stem and stem variantsWriting the question a learner reads first, and giving it alternative wordings.
- Choices and targetsThe per-item steps on choice, matching and ordering questions, and the ids that keep feedback pointing at the right thing.
- Hints: styles, unlock time and deductionsSix hint styles, when a hint becomes available, and what taking it costs.
- Solution steps and common errorsThe worked solution shown after an attempt, step by step, with its marks.
Learning
- Exit routing: outcome × confidenceWhich learning block runs after a right or wrong answer, and how sure the learner said they were.
- Active learning blocks: the eight techniquesWhat each technique asks a learner to do after they answer, and which are read-only.
- Building a learning block: before they choose, choices, after CheckThe segment tabs, the content slots, and the difference between the default reveal and tailored feedback.
- Conditional branches and placeholder tokensRefining which part of an answer went wrong, and dropping the learner's own answer into the feedback.
Review, publish and share
- Preview, new examples and render verificationSampling real versions of your question, and the check that blocks publishing.
- Teacher review: the full audit viewEverything in one page — stem, answer key, hints, timing, solution and learning content.
- Publish, unpublish and the completeness chipsWhat publishing actually changes, what unpublishing does not, and the seven chips that track readiness.
- Sharing and embedding one questionTeacher review links, student links, embed code, the token, and named links with their own grading.
- Share to Library, author credit and community scopePublishing a type to the community catalogue, what others can do with it, and how you are credited.
Building with AI
- AI build: one pass from setup to learningGenerating a whole question type at once, and the two things to read before you trust it.
- Assisted build: AI on every stepThe same editor as Manual build, with a Generate and Refine beside each section.
- The description, the AI brief and content-type preferencesWhat to write in the description, and the advanced brief that steers each section separately.
- Where the AI runs: copy, your GPU, your keyThree ways to run the same prompt, who pays for each, and what you need set up.
- Generating questions from a type, and batchesTurning a type into instances, previewing them before saving, and watching a batch run.
- Questions: instances, archive and deleteThe list of generated instances, what each row action does, and the difference between archiving and deleting.
Ladders
- Ladders: direction, not doseWhat a ladder decides, what it leaves to the delivery, and how to make one.
- The learning path canvasAdding question types, drawing paths, and pulling prerequisites in automatically.
- Step settingsPass score, how many questions, the predictive prompt, and what happens to a learner who already knows it.
- Path settings: gate modes, thresholds and if-not-readyWhat an arrow between two steps decides, and the five things that can happen to a learner who is stuck.
- Path profilesSix ready-made shapes for how a path behaves, and which suits which subject.
- Type groups on a ladderSeveral question types as alternatives at one step, rotating at random.
- Insert a ladder: copy in, or link as a blockReusing a whole ladder inside another, and the difference between a copy and a sealed link.
- Ladder options: retries, solutions, repeats and gradingThe advanced settings that apply to the whole ladder, including what happens when the question pool runs out.
- Confidence, remediation, pre-steps and interleavingThe four settings that apply to the whole ladder session rather than to one question type.
- Milestones, preview as a learner, and publishingCelebrating a step, testing the ladder as three different learners, and the checks before it goes live.
- The learning journey: skill status, this visit, review dueWhat the platform remembers about a learner across every ladder, and how it decides what they get next.
- Subjects, subject change and the dependency graphWhat a ladder's subject decides, what changing it does, and how dependencies become routes.
Deliveries
- Deliveries: the doseHanding a ladder to a class: the mode, the dates, the classes, and the readiness check.
- The eight delivery modes at a glanceWhat each mode does to the ladder, in one line each, and which to pick.
- Active learning modeThe ladder exactly as it is, with a due date — and the mode most homework should use.
- Assignment mode and the paperThe ladder's steps as a set paper, and the panel where you choose exactly what is on it.
- Exam modeTimed, one question per step, no feedback — plus reading time, extensions and the live monitor.
- Diagnostic modeProbing down the ladder until a learner stops coping, to find out where they actually are.
- Sprint modeFast rounds against a countdown, with lives, a leaderboard and adaptive weighting.
- Spaced review modeFilling each learner's own gaps from the repetition schedules, and why it needs a named learner.
- Live class and ChallengeRunning questions from the front of the room, and where the team round actually starts.
- How much of the ladder to serveTypes at each step, questions from each type, and how a group behaves.
- Links, classes, due dates and late policyThe four links a delivery carries, who it is assigned to, and what happens after the deadline.
- Results: release, export and the marking queueThe results table, what the attempt line counts, and when learners see their marks.
- What students see in a deliveryThe learner's screen: counters, hints, confidence prompts and what happens at the end.
- Duplicate, archive and delete a deliveryReusing one next term, taking one out of circulation, and why delete is sometimes refused.
Marking and the library