I am planning my first lesson
Start with a focused, student-ready resource and a teacher review gate.
Start here
Self-contained edition
This single file contains every chapter, the search index, styles, and search behavior. Save it locally or print it to PDF.
Type one or more words to search every chapter.
Use the path that matches what you need to do today. You can jump to any chapter from the full menu.
Start with a focused, student-ready resource and a teacher review gate.
Start here
Learn the interface: what every part of the screen does, how to generate, and how work is saved.
Start here
Prepare the room, route supports, recover calmly, and review evidence.
Start here
Start from a worked recipe and adapt it to your students and source.
Start here
Check learner experience, response options, and specialist workflows.
Start here
Set shared guardrails, coach teachers, and measure useful implementation signals.
Start here
Protect instructional time with a recovery sequence and a stable fallback.
Start here
AlloFlow is most useful when it helps you remove a specific barrier without changing the learning goal. Begin with one lesson you already teach, one trustworthy source, and one decision about what students need in order to participate. The platform can help create options; the teacher remains responsible for the content, rigor, accessibility, and way it is used.
For a first attempt, use Guided Mode and make one strong resource rather than exploring every tool. The process below usually fits into a planning period once you have your source ready.
The Launch Pad offers several starting points. The wording can vary slightly by deployment or interface language.
The complete-workspace card is labeled Full AlloFlow in the Launch Pad. Some interface text and older deployments still call the same route Full Platform; Guided Setup is another label you may see for the guided route.
Interface reference captured August 16, 2026 from the public AlloFlow deployment. Labels and available destinations can vary by deployment.
| Entry route | Use it when | Good first task |
|---|---|---|
| Guided Mode | You want a recommended sequence with prompts and checkpoints. | Prepare and deliver one differentiated lesson package. |
| Full AlloFlow | You already know the workspace or need to move among several tools. | Revise an existing project or build a custom workflow. |
| Learning Tools | You want to open a focused student-facing experience. | Explore a reading, writing, STEM, research, or SEL activity. |
| Educator Tools | You need a teacher, specialist, accessibility, reporting, or leadership workflow. | Review evidence, improve a document, or prepare professional materials. |
If the teacher role is protected in your deployment, use the school-provided access method. Do not share a teacher access code with students. You can change routes later, so the first choice is not permanent.
Have these four items ready:
Use the same source you would be comfortable projecting or handing to students. Remove names, student IDs, grades, disability information, behavior notes, health information, and other personally identifying details before pasting or uploading anything. See Privacy and responsible AI before using sensitive material.
Everything in the next section belongs in your planning time. By the time the first student sees anything, the resources should already exist, already be reviewed, and already be saved. Plan to open AlloFlow during class only to deliver what is finished.
Three reasons this matters more than it looks:
Finished before class starts:
Reasonable to do live: choosing which existing pathway a student gets, adjusting pacing, revealing an answer key you already checked, and opening a single tool link. If you find yourself generating new material while class is running, use the fallback instead and rebuild afterwards.
A practical habit: build the lesson at least a day ahead, then reopen it cold the morning of class. Loading your own project the way a student will meet it catches a surprising amount, and it costs two minutes.
From the Launch Pad, choose Guided Mode, then select the teacher role if prompted. Guided Mode organizes the work into a focused path that includes Source Material, Assignment Directions & Goals, Preview/Package/Deliver, and Review/Finish. Your deployment may offer additional optional steps.
If you lose your place, turn on Help Mode from the header or press the question-mark key. Help Mode explains supported controls and may show keyboard shortcuts. Press Escape to close an open dialog before trying another route.
Open Source Material and choose the route that fits what you have. Current workspaces may offer options such as opening a reading catalog, writing or pasting text, finding a resource online, generating from a topic, or importing a supported file.
Treat the source as the instructional ground truth for this lesson:
If you begin from a topic rather than a source, fact-check the resulting source before creating anything else from it. Generated material can sound confident while still being incomplete or wrong.
In Assignment Directions & Goals and the available setup controls, identify:
Describe barriers and supports, not diagnoses. For example, write “provide a short glossary and sentence starters” rather than including a student's name or disability label. Keep the intellectual target intact: simplifying access to a text is different from simplifying the idea students must understand.
Choose one output that directly serves the learning goal. Useful first choices include an adapted text, glossary, visual organizer, structured directions, short quiz, sentence frames, or another focused support.
Ask:
Do not create five versions simply because the platform can. A smaller set of purposeful options is easier for students to navigate and easier for you to verify.
Read every student-facing part. Check names, dates, links, calculations, answer choices, examples, translations, cultural references, and reading level. Confirm that the answer key, if present, matches the questions and is not visible in the student version.
Then check the experience:
Use the accessibility process in Accessibility and UDL for a fuller check.
Preview the actual delivery format whenever possible, not only the teacher workspace. In Guided Mode, continue to Preview/Package/Deliver. If you create a homework link, use Test latest student link before distributing its QR code or URL. For a live lesson, test the join path on a second browser profile or student device if your school setup allows it.
The preview should answer three questions:
If a feature depends on a microphone, audio output, AI provider, cloud service, or school filter, test it on the school network and device type students will use. Availability and persistence differ among hosted, desktop, embedded, and locally configured deployments.
Choose the delivery method that matches the lesson:
Before closing, use Save Project to download the project file when that option is available. Store it in an approved location with a useful name such as “ecosystems-food-webs-2026-08-13.” A project file is the safest way to resume in deployments where browser work does not persist after a tab closes. Reopen it through Load Project and confirm that the important resources return before relying on it for class.
Do not distribute a generated resource until you can answer yes to each item:
Generated content is a draft, even when it looks polished. Teacher approval is an instructional step, not an optional cleanup.
A strong AlloFlow lesson should feel like one coherent class task with useful choices, not a collection of unrelated apps. Students should know:
Avoid labeling pathways as “low,” “easy,” or “special.” Use neutral names tied to purpose, such as “audio and glossary,” “visual overview,” “practice with examples,” or “extension.” When possible, offer supports to the whole class while privately directing students to the options most likely to help.
You do not have to begin with a lesson at all. Every interactive STEAM Lab tool has its own direct web address that opens just that tool, with no sign-in for you or your students. Pasting one such link into whatever you already use (your LMS, a slide, a message home) is the lowest-commitment way to try AlloFlow, share it with a colleague, or give a family one good activity. When you are ready for more than one tool, come back to the fifteen-minute lesson above.
A direct tool link, captured August 16, 2026. The "AI extras: off" note in the header means no AI backend is set up on that device; the tool itself is fully usable either way. See Troubleshooting.
Protect instructional time first. Keep a stable fallback ready: the original source, a downloadable handout, or a simple discussion prompt. If a module is still loading, wait briefly, use its retry control if shown, or return to the previous view. If generation fails, preserve your source and directions before refreshing.
For a systematic recovery sequence, see Troubleshooting. For planning a fuller package, continue to Prepare a lesson. To teach it synchronously, continue to Live sessions.
This is the mechanical tour. Other chapters explain what to make and why; this one explains where everything is, which control does what, and how the three ways of generating work differ. If you have ever opened AlloFlow and not been sure what you were looking at, start here and read straight through with the app open beside you.
The first screen asks how you want to begin. It is not a settings page you have to get right, and nothing here is permanent.
Choose your workspace offers two doors into the same product:
Open a tool directly skips the workspace entirely:
Two more things live on this screen. Voice and device setup is optional and can be left alone. Top right, AI Backend Settings is where an AI connection is configured, and the globe control sets the language. The footer says it plainly: you can switch modes any time from the menu, so a wrong first choice costs you nothing.
Choosing Full Platform brings up two short gates. Both are one-time, and the second is skippable.
Who is using this device? Student, Teacher, Parent, or Independent Learner. This is what decides whether you see teacher controls at all, so pick Teacher on your own machine.
Quick Start then offers four steps of global context, beginning with grade level. Every step has a SKIP control, and skipping is a legitimate choice: everything Quick Start sets can be set later in Universal Settings. If you are exploring rather than building, skip it and look around first.
Once you are in, the screen is always the same four regions:
Nothing moves between these regions. When a chapter tells you to "open Universal Settings", it means the left column, every time.
The left column has two tabs at the top, Create and History. Create is where you work; History is what you have already made this session. And at the bottom right of the screen sits Student tools, which is how you look at the same work the way a student would.
Left to right:
The header has two states, and this catches people out. What is described above is the collapsed header. Expand it and a second, much larger bank appears: the App Language selector, Text and Voice controls, Translate, Documents, Assessment Center, and the AI, Tools, Learn and Bridge groups.
It also carries a row of icon buttons that are easy to overlook and worth learning, because three of them are the fastest routes to help: the lightbulb holds messages you may have missed, ? turns on Help Mode, the map starts the guided tour, the sparkles reopens setup and Guided Mode, and the cloud toggles Cloud Sync. The tour and setup icons appear in teacher mode only.
So if a control is missing from the header, it is almost certainly in the expanded state rather than gone. Collapse returns you to the compact strip and reclaims vertical space on a small laptop, and Maximize View gives the centre column the whole window.
There are two different language settings, they do different jobs, and confusing them is the single most common source of "why is this in the wrong language".
App Language is in the header. It changes the interface: buttons, menus, labels. Open it and you get a searchable list, your current language, and a Custom option if your language is not listed. Changing it may offer to regenerate the content you already have so it matches, and it warns you first, because unsaved changes to the current text can be lost in that regeneration.
Output and translation languages live in Universal Settings, not the header. These decide what language the resources you generate come out in, and which translations get attached. A glossary panel will tell you directly when none are set, with a line reading that no translation languages are set and pointing you to Universal Settings.
So: header for the language you read the app in, Universal Settings for the language students receive.
The header line reads Apply to most resources you generate, and the collapsed state summarises itself, for example 3rd Grade · English. That summary is worth glancing at before every generation.
Two ways to fill it in:
Inside are grade level, output language and translations, whether to use emoji for visual support, a Differentiation Set (for instance Target Level Only), and a default Image Style used by Visuals, Glossary, Timeline and Concept Sort unless a tool sets its own.
The detail that makes this panel trustworthy is the small coverage note under many settings, reading something like Applies to 12 of 19 resource types. AlloFlow is telling you exactly how far a setting reaches rather than implying it governs everything.
These apply to new work only. Set them before you generate, not after.
Below the settings is the tool finder, and it has three ways to narrow a long list:
The list underneath holds the resource tools: Analyze Source Material, Glossary, Text Adaptation, Word Sounds, Visual Organizer, Note-Taking Templates, Anchor Chart, Lesson Images, FAQ Generator, Writing Scaffolds, Activities, Interview Mode, Sequence Builder, Concept Sort, Document Analysis, Assess, Lesson Plan, Assignment Directions, and the entry points to the STEAM Lab and Adventure Mode.
The centre column begins at Source Material, because most resources are generated from something. There are three routes, offered as cards:
The same routes exist as a row of buttons on the Source Material card in the left column: Upload, Load Project, Link, Generate, and Books. Under them is a paste box that accepts curriculum text directly, or a PDF, image, text or audio file up to 20MB.
Once a source is in, Analyze Source Material reads it back to you, so you can check the AI understood it before you build eight resources on top of a misreading.
You do not always need a source. Some tools invent their own material. But when a tool warns that a step produced nothing, a missing source is the usual reason.
Click a tool in the left column and its panel expands in place. Every tool panel has the same shape:
Change tool swaps the open panel without collapsing everything and starting again.
The full pack is the "do it all" path. It generates a coordinated set of resources from one source in a single run.
It lives at the very bottom of the tool list, under an Auto-Configure card, which is easy to scroll past.
The full pack is the fastest route from one text to a usable lesson set. It is also the easiest way to spend a lot of generations at once, which is exactly why the planner shows the estimate first.
This is the most powerful path in the app and the least discoverable, because it lives inside AlloBot rather than on a button of its own.
How to turn it on. Open AI Guide and Assistant from the header to bring up the AlloBot panel on the right. Near the bottom, tick Auto-Fill Settings, marked Recommended. That toggle is what puts AlloBot into Blueprint Mode. Then describe what you are teaching in ordinary language.
What happens next. AlloBot tells you it is analysing context and designing a lesson blueprint, and it says up front that you will be able to review and change the plan before anything is generated. It then presents a Lesson Blueprint: an ordered list of the resources it proposes to build.
Reviewing and changing the plan. Nothing is generated yet, and this is the point of the whole feature.
Building it. Generate Plan starts the run. A progress line reads Building, N of M steps finished. If you need to stop, Stop after this step finishes the step in progress and then halts: everything already finished is kept, and the remaining steps show a Rebuild control so you can run them one at a time later.
Two things worth knowing before you start. If there is no source text yet, AlloFlow warns that some resources may not generate and suggests adding or generating a source first. And previous plans are archived rather than discarded, so an earlier blueprint can be restored, though not while a plan is still generating.
Two shortcuts do more work than anything else in this chapter.
Ctrl+K (Cmd+K on a Mac, or Ctrl+Shift+P) opens the command palette. Type what you want in ordinary words and it takes you there. It reports how many commands match, you can star the ones you use constantly, and Esc closes it.
? turns on Help Mode. With it on, click any button, panel or tool and AlloFlow explains that specific element in plain language. This is the fastest way to answer "what is this control" without leaving the screen you are on, and pressing Esc turns it back off.
Work lives on the device, so saving is something you do deliberately.
.json extension. This is your backup and the way to move a lesson to another machine.If a lesson matters, save the project. Browser data clearing removes on-device work, and the project file is what survives it. Saving, loading, and managing storage covers the whole picture, including the recovery key and how to stop AlloFlow running out of room.
You will sometimes see a line on a tool panel reading Needs AI setup, offering to connect an AI or to use AlloFlow inside Gemini Canvas. Nothing is broken. AlloFlow separates the things that need a model from the things that do not.
Generation needs AI. Browsing tools, the STEAM Lab simulations, Universal Settings, saving and exporting, and everything you have already generated all keep working without it. If you are evaluating AlloFlow before your district has chosen a model, you can still see most of the product.
For the three ways to connect one, see Troubleshooting; for what leaves the device when you do, see Privacy and responsible AI.
Now that the screen makes sense, Prepare a lesson shows how to use it well, Universal Settings goes deeper on the settings panel, and AlloBot covers talking to the assistant beyond Blueprint Mode. Keep Quick reference beside you until the shortcuts are habit.
AlloFlow has more settings than any one teacher needs, which is a problem only if nobody tells you which ones matter. This chapter sorts them into the two kinds that behave differently, then covers the three ways to get help without leaving the screen you are on.
The distinction that prevents most confusion:
If output came out wrong, look at Universal Settings. If the app itself is behaving unexpectedly, look at app settings.
Open it from the top of the left column. Collapsed, it summarises itself, for example 3rd Grade · English, and that one line is worth reading before every generation.
Two ways to fill it in. AI Match infers the settings from your source material and goal. Manual puts you in control of each one. AI Match is a good starting point; Manual is what you want once you know your class.
What is inside:
The coverage notes are the important detail. Under many settings sits a small line reading something like Applies to 12 of 19 resource types. AlloFlow is telling you exactly how far that setting reaches instead of implying it governs everything. When a setting seems not to have worked, check whether the tool you used is inside its coverage.
They apply to new work only. Changing grade level does not rewrite what you already made. Set them first, then generate. Universal Settings goes further on using them well.
AI Backend Settings is on the Launch Pad, top right, and again under AI in the header's More information menu, which also holds model diagnostics and a usage meter.
Until an AI is connected, tool panels show a line reading Needs AI setup, offering to connect one or to use AlloFlow inside Gemini Canvas. Nothing else is broken by this: browsing, the STEAM Lab, settings, saving and exporting all work. Only generation waits.
The usage meter is worth knowing about before you run a full pack, because a pack is the fastest way to spend a daily quota.
App Language changes the buttons, menus and labels. It sits in the expanded header under its own APP LANGUAGE label, with a dropdown for the listed languages and an Enter Language box beside it for one that is not listed.
Changing it may offer to regenerate the content you already have so it matches, and it warns you first, because unsaved changes to the current text can be lost in that regeneration. Say no if you have unsaved work you care about, then change it again once you have saved.
Remember the pairing: App Language for the language you read the interface in, Universal Settings for the language students receive. Setting one does not set the other.
Offered on the Launch Pad as Voice and device setup, and marked Optional because it is. This is where microphone and voice output get configured for AlloBot's Talk mode and read-aloud.
The models that power it are downloaded once and kept on the device, so speech works without sending audio anywhere. They are also the largest thing AlloFlow stores. Saving, loading, and managing storage covers the size and how to reclaim it.
On first run AlloFlow asks whether this is a Student, Teacher, Parent, or Independent Learner. That answer decides whether teacher controls appear at all.
The header breadcrumb always shows the current answer, reading something like TEACHER / SOURCE MATERIAL, so a glance confirms you are not accidentally in a student view. Student tools, at the bottom right, is the deliberate way to look at your work as a student would.
You can also move between Guided Mode and Full Platform whenever you like. The Launch Pad says so directly: you can switch modes any time from the menu.
A toggle in the More information menu. Off by default, which is consistent with the rest of the product: work stays on the device unless you choose otherwise.
How much of the device AlloFlow may use, which models are downloaded, and how to recover a previous session. All of it is in Saving, loading, and managing storage.
Help Mode is the one to learn first. Turn it on and click any button, panel or tool, and AlloFlow explains that specific element in plain language. Turn it off and everything returns to normal, dismissing the tooltips and spotlights.
Two ways in:
Esc turns it off. This is the fastest possible answer to "what is this control", because it never takes you away from the thing you were doing.
The tour spotlights each part of the interface in turn, with an explanation of what it is for: the input panel, the accessibility upload, the AI Guide, the tool finder, Universal Settings, source analysis, and then each tool in the list.
How to start it. Three routes, and the first is the one to remember:
One deliberate touch worth knowing: while the tour is running, every tool is shown, with your purpose filters set aside, so you see the whole set rather than whatever subset was filtered when you started.
The remediation pipeline has a tour of its own, reached the same way from inside that tool.
Ctrl+K (Cmd+K, or Ctrl+Shift+P) opens it. Type in ordinary words and it finds the command.
Three things make it more useful than a search box:
Esc closes it. Between the palette, Help Mode, and the tour, you should rarely need to hunt through menus.
That is the whole configuration surface that matters on day one. Everything else can wait until you meet it.
AlloFlow keeps your work on the device rather than in an account. That is what makes it usable without sign-in, rostering, or a district contract, and it is also the thing most likely to lose you a lesson if nobody explains it. This chapter explains where your work actually lives, how to save and reload it, and how to keep the device from filling up.
Read Privacy and responsible AI for why the design is this way. This chapter is the operating manual.
Everything you generate is held in the browser's own storage on the machine you are using. Nothing is uploaded, which has three practical consequences worth stating plainly:
The fix for all three is the same, and it is the next section.
Save Project writes your entire session to a single file: the source material, every resource you generated, and the settings that produced them.
.json extension itself.This is the one habit worth building. If a lesson took you more than a few minutes to build, save the project. The file is device-independent, survives a browser wipe, and is how you move work between home and school.
In Gemini Canvas, this matters even more. Nothing survives closing the tab except files you have downloaded. The project file is your save.
Saving a project preserves your ability to keep working. Exporting produces the thing students or colleagues actually receive. They are different jobs and you usually want both.
The Export options include:
For the full treatment of formats, margins, and what prints well, see Documents and printing.
AlloFlow has a Storage and recovery panel, and on a device used all year you will eventually want it.
If the app cannot find your previous session it says so directly, reporting that no restorable workspace was found rather than opening silently empty. You can also choose to work without device recovery, which is the right choice on a shared or public machine where you do not want work persisting after you walk away.
Where a recovery key is offered, it is shown once. Write it down at the moment it appears, because it cannot be shown again.
Because browser storage is finite, AlloFlow lets you choose how much of it to use:
| Preset | What it targets |
|---|---|
| Standard | About 20 workspaces, 150MB, and 50 offline resources. This is the normal behaviour. |
| Compact | About 4 workspaces, 50MB, and 20 offline resources. Older unpinned draft-only work may expire. |
| Automatic | Uses Standard normally, and drops to Compact when the device reports storage pressure. |
Automatic is the sensible default for a school device. Choose Compact deliberately on a Chromebook that is short of space, and understand what you are agreeing to: unpinned drafts you have not saved as projects can be dropped. Anything you have saved as a project file is unaffected, because that file is outside the browser.
If you ever see a message that a plan could not be archived because storage may be full, this panel is where you go.
Voice features download their models once and then keep them on the device, so speech works without sending audio anywhere.
The panel shows each model's download size, whether it is already on this device, and the total model cache. These are the largest single thing AlloFlow stores, so if you need space back and you do not use voice features, this is the first place to look.
If you have run a document through the accessibility remediation pipeline, the result is held on the device so you can reopen and review it. The panel offers to open those results directly. See Make a document accessible for that workflow.
Check these in order:
If none of those apply, Troubleshooting covers the wider recovery sequence. And if the answer turns out to be that no project file was ever saved, that is the habit worth changing rather than a fault to chase.
Everything on this page is a reminder, not a lesson. If a line raises a question, the linked chapter answers it. Print this one for pilot teachers.
| Press | What happens |
|---|---|
| Ctrl+K (or Cmd+K) | The command palette: type what you want ("open the glossary", "math minute") and it takes you there. Ctrl+Shift+P does the same. |
| ? | Help Mode on or off: click anything highlighted for a plain-language explanation. |
| Esc | Closes the palette, leaves Help Mode, backs out of most dialogs. |
| You want | Go to |
|---|---|
| Narrow the tool list | Find a tool at the top of the sidebar (says in words when a filter is on) |
| A message that vanished | The header lightbulb → Messages |
| Your saved work | AI settings → Open saved work |
| A finished copy vs a blank copy | Export: Print / Save as PDF vs Worksheet (Documents and printing) |
| One tool to share with anyone | Its direct link, no sign-in needed (Start here) |
| Timed math practice | Math tool → Mode, or Ctrl+K then "math minute" (Math Fluency) |
| Turn Adventure off for an assignment | Project Settings → "Include Adventure in this assignment" (Adventure Mode) |
| "AI extras: off" in the STEAM Lab | Nothing is broken; click the pill for the three ways to turn AI on (Troubleshooting) |
| Fix an inherited PDF | The remediation workflow: audit, Make Accessible, review, export with evidence (Make a document accessible) |
This chapter turns the quick start into a repeatable planning routine. The aim is not to generate the largest collection of resources. It is to build a small, coherent package in which each item has a clear job: establish meaning, remove a barrier, support practice, or reveal learning.
Begin with the learning goal and evidence of success. Then choose AlloFlow tools because they serve that design. If you are preparing your first resource, start with Start here.
Write the shared target before selecting formats:
Students will explain how energy moves through a food web and support the explanation with evidence from the source.
Next, identify where access may break down. A barrier is a feature of the task or environment, not a label attached to a student.
| Possible barrier | Planning response | Possible resource |
|---|---|---|
| Dense background text | Clarify structure and foreground essential ideas. | Chunked or adapted text with preserved key concepts |
| Unfamiliar academic vocabulary | Preteach a small set of terms in context. | Glossary, examples, visuals, or word practice |
| Multi-step directions | Make sequence and completion criteria visible. | Numbered directions, checklist, or visual organizer |
| Limited background knowledge | Build a bridge without replacing the core source. | Brief overview, timeline, concept map, or worked example |
| Language demands obscure content knowledge | Support comprehension and expression. | Translation, bilingual glossary, sentence frames, or oral option |
| One response mode excludes participation | Offer another way to demonstrate the same target. | Written, spoken, visual, organized, or interactive response |
| Students are ready for greater complexity | Extend the reasoning rather than add busywork. | Competing evidence, transfer task, debate, or higher-depth prompt |
Do not assume every student needs every support. Too many choices can become a new barrier.
Open Source Material in Guided Mode or the full workspace. Use a source you are permitted to share and that you can evaluate. Depending on the deployment, you may be able to paste or write text, import a supported file, open a reading catalog, use a web resource, or generate a starting text from a topic.
Interface reference captured August 13, 2026 from the public AlloFlow deployment. Availability and labels vary by deployment.
Before generation, make the source self-contained:
AlloFlow can transform a weak source into polished-looking weak material. Source quality is therefore one of the most important teacher decisions in the workflow.
If an analysis or source-review step is available, use it to notice complexity, vocabulary, organization, possible misconceptions, and missing context. Treat automated analysis as a planning suggestion, not a measurement of a student's ability or a final judgment about the text.
Create a short “must preserve” list before adapting:
This list protects rigor. A clarified version may shorten sentences or explain terms, but it should not remove the causal relationship students must reason about.
In Assignment Directions & Goals, write what a student needs when the teacher is not standing beside the screen:
Use concrete verbs such as compare, explain, model, revise, justify, or solve. “Complete the activity” does not tell students what thinking matters.
Most of these live in one place. Set the grade level, output language, translations, standards, interests, and Depth of Knowledge in Universal Settings before you generate, so every part of the pack agrees. Set subject and response format where the tool offers them. Do not enter a student name or confidential profile to personalize a resource. Describe a support in general terms, such as:
Interests can increase relevance, but they should not stereotype students or distort the content. A theme is useful only if it makes the task clearer or more engaging.
The sidebar's Find a tool panel narrows the tool list by purpose (Recommended, Make accessible, Engage, Assess and deliver) or by search. It filters what you see; the tool cards below it are what actually create. While a filter is on, the panel says so in words ("A filter is on: showing 8 of 22 tools"), and you can only hide the panel when no filter is active, so tools can never go missing without an explanation on screen.
Interface reference captured August 16, 2026 from the public AlloFlow deployment.
A practical package usually has three to five parts:
You do not need a different tool for every part. A well-designed source, clear directions, and one strong response task may be enough.
| Output | Best use | Teacher check |
|---|---|---|
| Adapted text | Reduce unnecessary language load while preserving the concept. | Compare line by line for omissions, changed claims, and lost evidence. |
| Glossary or word support | Prepare students for a small set of essential terms. | Check definitions in context, pronunciation, translations, and examples. |
| Visual organizer | Show relationships, sequence, categories, or cause and effect. | Confirm reading order, labels, text alternatives, and logical connections. |
| Structured directions | Make a complex task actionable. | Test the steps without relying on verbal explanation. |
| Sentence frames or writing scaffold | Support entry into disciplinary expression. | Make frames optional or gradually removable; do not write the reasoning for students. |
| Quiz or quick check | Reveal current understanding efficiently. | Verify answer keys, distractors, wording, scoring, and alignment to the goal. |
| Concept sort or sequence | Make student thinking visible through classification or order. | Confirm that categories are defensible and ambiguity is intentional. |
| Adventure, game, or simulation | Create repeated practice, exploration, or a meaningful context. | Check that play advances the target rather than hiding it. |
| Audio or read-aloud support | Provide another route into text. | Listen for pronunciation, pacing, language, and missing content. |
| Extension prompt | Increase transfer, argument, design, or synthesis. | Extend cognitive demand instead of adding more of the same work. |
Generate the anchor resource first. Compare it to the source and learning goal before creating dependent items. A quiz made from an inaccurate adaptation will repeat the error; a translation of that quiz will spread it further.
Use a simple review loop:
Prefer targeted revision requests such as “restore the explanation of evaporation in paragraph three” over “make it better.” Keep the original source available for comparison.
Every teacher should check facts, attribution, age appropriateness, and alignment. Also use a subject-specific pass:
Never rely on a plausible explanation without working through it yourself.
Students may use different representations or response modes while working toward the same essential understanding. For example:
If a pathway changes the goal, say so explicitly and base that decision on the student's instructional plan and teacher judgment, not an automated recommendation.
Use neutral, useful labels:
Avoid public labels based on perceived ability. In a live lesson, use private individual or group routing when that better protects student dignity. See Live sessions.
When appropriate, let students choose a support and switch if it is not helping. Teach them to explain the choice: “I used the visual organizer because I needed to compare two causes.” This makes accessibility a normal learning skill rather than a hidden accommodation.
Open the resources in the order students will encounter them. Verify:
For a 45-minute middle-school lesson, a realistic sequence might be:
Adjust for your class; the sequence is a planning example, not a built-in timing promise.
Continue to Preview/Package/Deliver in Guided Mode or use the available preview and export controls in the full workspace.
Two export formats look alike and are not: Print / Save as PDF produces a finished copy to read, while Worksheet produces a blank copy to write on, with ruled answer lines, fill-in bubbles, and a Name and Date header. The export screen explains each format under its selector.
If your lesson has both an adapted text and a glossary, the Worksheet format also offers a fill-in-the-blank version: glossary terms in the passage become numbered blanks with a word bank, and the teacher copy carries the answer key. It works in any lesson language.
Test again after a meaningful revision. Previewing an earlier version does not verify the final package.
Use Save Project before leaving the workspace when that option is available, then test Load Project with a copy. Keep final exports separate from working files. A simple naming pattern makes team use easier:
Store files according to district policy. Do not put student responses or identifying notes in a shared curriculum folder.
For independent delivery, include a short message outside AlloFlow that states the goal, due date, expected product, accessibility contact, and fallback route. For a colleague, include the source, teacher-approved resources, answer key, intended sequence, and any deployment requirements.
Continue with Accessibility and UDL for a deeper learner-experience review, Live sessions for synchronous delivery, or Review evidence and plan next steps for using classroom evidence responsibly.
Universal Settings is one card that holds the choices every generator inherits. Set the grade level once and the glossary, the quiz, and the adapted text all target that grade. Set it per tool instead and the parts of one lesson pack quietly disagree with each other.
Open it before you generate anything. That single habit prevents most of the "why does this pack feel mismatched" problems teachers report.
If you have not made a resource yet, start with Start here. To see how these settings fit a full planning routine, see Prepare a lesson.
The panel gathers the settings that would otherwise repeat in every tool.
Interface reference captured August 16, 2026 from the public AlloFlow deployment. The note at the top and the "applies to" lines under each control state exactly which resources a setting reaches.
| Setting | What it does | Where you notice it |
|---|---|---|
| Grade level | Sets the reading and complexity target for new resources. | Sentence length, vocabulary, and question difficulty. |
| Output language | Sets the language generated content is written in. | The text students read. |
| Translations | Decides whether a second-language version is attached. | A translation block beside the main content. |
| Standards | Supplies the standard or framework the content should align to. | Quiz items, objectives, and lesson plans. |
| Student interests | Adds a theme or context to examples. | Word problems, scenarios, and prompts. |
| Depth of Knowledge | Sets the level of thinking the task should demand. | Question stems and task verbs. |
| Emoji | Adds emoji as visual cues in generated material. | Glossary terms, headings, and directions. |
Each control tells you how many kinds of resource actually use it. Read that number before you spend time on a setting. Some controls reach almost everything; others reach two or three tools.
You can collapse the card once you are done. The summary line stays visible, so you can check your choices without opening the panel again.
Universal Settings applies to new work only.
A resource keeps the settings it was built with. If you generate a glossary at grade 3, then change the grade to grade 7, the glossary you already have is still grade 3. Nothing rewrites itself.
So if a pack looks inconsistent:
This is worth checking before you conclude a tool got something wrong. A mismatched pack is usually a settings history problem, not a generation problem.
The grade level is a target, not a guarantee. Generated text often lands above the grade you asked for, and asking for more research or more detail tends to push it higher still.
Do this before you share:
If it reads high, ask for a lower grade than you want and check again. Set the grade for the students who need the most access, then offer the original source to students who are ready for it. Lowering the reading load is not the same as lowering the learning goal. See Prepare a lesson for how to protect the intellectual target while changing the reading level.
Two separate controls do two different jobs.
Output language is the language the resource is written in. If your class works in Spanish, set it to Spanish and the content comes back in Spanish.
Translations decides whether a second version rides along with it. It offers three kinds of answer:
A hint line under the control always says in plain words what will happen, such as "Resources in Spanish will also include an English version." You do not have to open the list to know where you stand.
The Translations control appears only when it would do something. If your interface and your output language are both English, there is nothing to translate into, so the control stays hidden. It appears as soon as the two differ.
Two cautions:
Emoji make headings and glossary terms easier to scan for some students. They also carry costs worth knowing about:
Turn emoji on when the visual cue helps and you have previewed the result. Turn them off when the same resource will be printed, read aloud, or used in a spelling or letter puzzle.
Before you generate the first resource for a lesson:
Then generate one resource and read it before making anything else. See Accessibility and UDL for the learner-experience check, and Troubleshooting if a setting does not seem to take effect.
A Live Session lets a teacher pace a lesson, route resources to the class, groups, or individuals, run checks for understanding, and respond to student signals. It works best when the lesson has been prepared and previewed before students join.
Live availability, join URLs, storage, and network behavior depend on the school's deployment. Rehearse on the same network and device types students will use. Always keep a non-live fallback ready.
Use a Live Session when you need:
Use a homework link, accessible document, learning-management assignment, or other asynchronous route when students need flexible timing, the connection is unreliable, or the work should not depend on everyone being online at once.
The Live Session Center can toggle between two pacing approaches.
| Mode | What it supports | Teacher responsibility |
|---|---|---|
| Teacher-paced | The class follows the teacher's current resource; useful for modeling, discussion, and coordinated transitions. | Announce each change, watch delivery status, and allow enough time for assistive technology and reading. |
| Student-paced | Students work more independently while the teacher monitors and may route targeted resources. | Make sequence and finish criteria explicit; do not assume every student is on the same screen. |
Switch deliberately. A whole-class presentation or follow-up may require Teacher-paced mode, while independent work may be less disruptive in Student-paced mode. Tell students before changing the mode so a screen transition does not feel like lost work.
Complete the workflow in Prepare a lesson. Keep the live resource set small enough to navigate while teaching. A useful sequence is:
Open every student-facing resource once. Remove teacher notes and answer keys. If the workspace marks an item as teacher-only, do not try to send it to students.
Before class:
If school filters, browser privacy controls, embedded environments, or blocked real-time services interfere, use the approved deployment guidance. Do not ask students to disable device security settings.
Decide how students will identify themselves. Follow district policy and the purpose of the lesson:
Prepare groups before the activity when possible. Grouping should serve an instructional purpose and can change during the lesson.
Write the fallback task where students can see it. Decide whether the session code may be projected and hide any roster or teacher-only material before sharing your screen. Have headphones available for audio supports when possible, and avoid starting every student's read-aloud at the same time.
Use the available Live Session or class-session control in the header or delivery workflow. Choose the session options shown in your deployment. Once a session is active, the Live Session Center becomes available to the teacher.
Open the session-code or projection view and verify that it shows only information students should see. Keep the teacher workspace on a non-projected screen if possible.
Students open the school's approved AlloFlow student link, choose the student role if prompted, and enter the session code and approved name or codename. The precise sequence can vary by deployment.
Ask students to stop after they reach the waiting or first-resource screen. This creates a clean point for checking the roster before instruction begins.
In the Live Session Center, compare the visible roster with the students who should be present. Look for duplicate or unexpected codenames. A connected indicator or recent check-in is useful operational information, but it does not prove attention or understanding.
If an unknown participant appears, pause before sharing material. Use the session controls available to remove or resolve the entry, and change the session if required by school procedure.
Open the first student-facing resource and send or present it. Ask students for a simple confirmation, then inspect delivery status. Resolve a failed or pending device before launching a timed activity.
The center groups the main controls under Run, Guide, and Signals. Labels and available activities may vary with loaded modules and deployment settings.
Use Run for student interactions and the prepared lesson sequence. Current workspaces may include:
Launch one activity at a time. State the purpose, response expectations, time limit, and what students should do after submitting. Close or conclude an activity before moving to a different response format.
Use Guide to manage the flow around the activities. Depending on the current resource and loaded tools, it may provide:
Presenter cues are private teacher reminders, not student directions. Keep anything students must know in the student-facing resource or say it aloud and provide it in another accessible form.
Students can use preset signals to communicate a need without interrupting the whole class. The Signals area shows the student's codename and selected phrase. Acknowledge the signal, respond privately when appropriate, and clear it when addressed.
A signal is a request for support, not a diagnosis or behavior score. Teach a brief routine before the lesson:
Do not leave the signal panel projected, especially when only one student is signaling.
The live lesson run controls can route a student-facing resource to the whole class, a group, an individual student, or a selected set when those options are available.
Use whole-class delivery for the shared core. Use targeted delivery for a support, catch-up resource, alternate representation, or extension. A targeted route should not publicly announce why a student received it.
Before sending:
The center may show whether a learner has a target resource, whether it is loading, pending, open, or failed, and whether the device is connected or quiet. Treat these as troubleshooting signals:
Status does not show whether the student read, understood, or completed the resource. Ask for evidence rather than inferring learning from a green indicator.
Individual and group targets can take precedence over the whole-class resource. If a student appears “stuck” on a different item, check for an individual or group assignment and use the available release control before sending the class resource again.
In Teacher-paced mode, changing the current class resource can move connected student devices and may affect devices that reconnect while it remains current. Give a spoken and visible transition cue first.
In Student-paced mode, a whole-class follow pointer may not behave like a synchronized presentation. Use explicit directions and targeted sends as needed, then verify the result on the student device. If a prepared whole-class follow-up reports that Teacher-paced mode is required, switch modes only after warning the class.
Use Quick Check for a decision you are prepared to make. Ask one focused question such as “How ready are you to explain the relationship?” Display what each response option means, allow a short wait, and decide in advance what will happen at each pattern of responses.
Do not use confidence as a substitute for knowledge. Pair it with a prompt, example, or subsequent evidence task when accuracy matters.
Use Word Cloud for short, non-sensitive contributions: a key term, observation, prediction, or theme. Tell students not to enter names or private stories. Review or moderate responses before displaying them when the interface offers that control. A word's visual prominence represents frequency, not importance or correctness.
Give a bounded prompt, response length, and success criterion. If the platform can generate feedback, review the selected response and generated feedback before treating it as instructional guidance. Do not submit identifying text or confidential student work to an AI provider.
Use drawing to reveal relationships, models, processes, or vocabulary, not artistic talent. Provide a non-drawing response option when motor, visual, device, or cultural factors make the activity inaccessible. Remind students not to draw identifying or inappropriate content.
Verify each item and answer key before class. Start with an untimed practice item if students are new to the interface. A missing or unscored response can reflect connection trouble, not lack of knowledge; check the delivery state before drawing conclusions.
Enable live questions only when you can actively moderate them. Establish norms for relevance and privacy. Review submissions before sharing them with the class, and use the school's normal safety process for a message that indicates urgent risk.
Activity Pulse and activity snapshots can help you see counts, statuses, or response patterns while teaching. Use them to choose a next move:
These indicators are incomplete classroom evidence. They may omit response content, depend on the connection, or represent device events rather than learning. Never turn a live status directly into a grade, diagnosis, disciplinary judgment, or high-stakes placement decision.
Check the join URL, session code, role, and codename entry. Confirm the teacher session is still active. Have the student reopen the approved link rather than using another student's page. If joining still fails, give the fallback resource and continue the lesson.
Confirm the student is connected, the resource is student-facing, and the correct audience was selected. Check for a conflicting individual or group target. Use the center's delivery state or session-health control, then try a targeted resend once. If it still fails, share the fallback format.
Pause transitions. State the title of the screen everyone should have. Check the pacing mode and current class resource, then inspect individual or group assignments. Release outdated targets when appropriate. Avoid creating a second session until you understand whether the original is recoverable; two active codes create more confusion.
Close the activity panel if possible and return to the core resource. Capture the instructional question orally, on paper, or in a simple response form. Do not spend the lesson repeatedly reconnecting an optional activity.
Keep the tab and session code available while devices reconnect. Do not assume a quiet presence indicator means a student intentionally left. Move to the posted fallback if the connection is not restored quickly enough for the learning goal.
See Troubleshooting for a fuller diagnostic sequence.
Do not treat closing a tab as the same as ending a class session. Use the session's End Session control. If an end-session preview is shown, review what will close, what evidence or follow-up is available, and whether any student is still active before confirming.
After ending:
The end-session summary is a starting point for teacher reflection, not a complete record of learning. Continue with Review evidence and plan next steps.
Before students join:
While teaching:
Before ending:
For a detailed learner-experience check before the next session, see Accessibility and UDL. For privacy boundaries, see Privacy and responsible AI. If your lesson includes a story mode, Adventure Mode covers when students see it and how resume stays scoped to this lesson.
Adventure Mode turns a lesson into a short interactive story. Students make choices, answer questions drawn from your source material, and earn XP as they go. It is optional, and as of the August 2026 update it only appears when you want it to.
If you have not built a lesson yet, start with Prepare a lesson.
Two things must both be true before the Adventure panel appears for students:
| Condition | Where you control it |
|---|---|
| You left Adventure on for this assignment. | Project Settings has an "Include Adventure in this assignment" switch. Turn it off and the panel disappears from the student view entirely. |
| There is a lesson to build a story from. | Adventure needs source material or an analysis to work with. An empty workspace does not advertise it. |
The switch defaults to on, so lessons you shared before this update behave exactly as they did.
A student's saved adventure belongs to the lesson it was made from. If they open a different lesson, they are not offered "Resume Adventure" from last week's story, and choosing an old save from a different lesson politely refuses rather than pulling them out of today's work. This closes a real classroom problem: students quietly resuming a past adventure instead of attending to the current lesson.
One honest note: adventures saved before this update carry no lesson tag, so the very first resume after the update may still offer an older story. Every save made from now on is tagged.
Adventure has its own language control with three settings: the lesson language only, the lesson language with a translation, or a multilingual mix. When a translation is included, the second language now follows your Universal Settings translation choice instead of always being English. See Universal Settings for how that control works.
Family mode is the parent-facing way to use AlloFlow: a parent or caregiver authoring support for their own child. It was reviewed and tightened in the August 2026 update, so what a parent can reach is now a decision, not an accident.
If you are a teacher sending work home instead, see Prepare a lesson; this chapter is about a family running AlloFlow themselves.
Family mode is the authoring experience, with home-appropriate language. A parent can build adapted texts, glossaries, quizzes, visuals, and activities for their child, use the STEAM Lab and Learning Hub tools, and manage everything saved on their own device. Several tools introduce themselves in family terms: the glossary presents as a word helper, the lesson plan as a family guide, and the alignment tools in plain language.
Two things are deliberately kept:
These are school-role surfaces, closed to family mode by design:
Universal Design for Learning is a planning approach, not a button or a collection of “special” versions. Start with a meaningful shared goal, anticipate barriers in the task and environment, and provide useful ways to engage with information and demonstrate learning. Then test what students will actually experience.
AlloFlow can help create accessible options, but it does not certify that a lesson, document, website, or activity meets a student's needs or a legal standard. Generated text, translations, images, audio, captions, interaction patterns, and exports require teacher review and, when appropriate, specialist testing.
Separate the learning target from the method used to reach it.
Write what must remain constant and what can vary. This prevents a support from accidentally removing the skill students are meant to practice.
Ask where the lesson depends unnecessarily on one kind of perception, language, movement, memory, timing, or social participation.
| UDL dimension | Planning question | Examples of flexible support |
|---|---|---|
| Engagement | How can students understand the purpose, enter the task, sustain effort, and ask for help? | Relevant choice, clear goal, predictable sequence, timer, progress cue, partner or independent route |
| Representation | How can students perceive and make meaning from the essential information? | Read-aloud, glossary, translation, visual model, captions, transcript, clarified layout, worked example |
| Action and expression | How can students navigate, practice, organize, and show the target? | Keyboard access, speech or typing, drawing with a text alternative, organizer, sentence frame, oral or written response |
Choice is not automatically accessible. Every offered option must work, and students need enough information to choose without becoming overwhelmed.
After the lesson, ask which supports students selected, where they stopped, what still required teacher rescue, and whether each response mode revealed the intended learning. Use that evidence to improve the next version. Do not infer a disability, preference, or permanent “learning style” from one interaction.
An accessible lesson should work before a student opens optional tools.
The opening screen or page should state:
Keep required content in the main reading and interaction order. Do not hide essential directions only in a tooltip, image, teacher speech, color cue, or optional audio.
Use a short sequence with consistent names. Mark optional supports by purpose, such as Listen, Vocabulary, Visual overview, or Planning organizer. Avoid sending students back and forth among many tools for one simple task.
When a lesson has several resources, include a checklist or numbered route. If students work asynchronously, add an explicit stopping point and a way to resume.
Prefer short, concrete directions while preserving disciplinary vocabulary students need to learn. Define new terms in context. Break long procedures into steps and pair icons with text labels rather than relying on icons alone.
Plain language is not the same as lowered expectations. Students can reason about complex ideas through clear instructions.
The exact supports and labels available depend on the current view, loaded modules, browser, and deployment. Turn on Help Mode or press the question-mark key to inspect supported controls. Do not promise a feature to students until you test it in the delivered resource.
Text Settings can provide controls such as text size, spacing, font, reading theme, or color support. Some views or exports may also offer light, dark, sepia, warm, cool, or high-contrast themes, a reading ruler, line focus, or reduced-animation options.
Use these controls to increase perceptibility, but remember:
Never encode “correct,” “group,” “urgent,” or “selected” through color alone. Add a label, symbol, pattern, or position that carries the same meaning.
Depending on the view, read-aloud may appear as Read This Page, Read Aloud, or an Immersive Reader-style experience. Voice Settings may allow voice, speed, and volume changes. Some resources may also offer highlighting, focus reading, a reading ruler, word or picture support, syllable cues, or other reading tools.
Before assigning:
Synthetic speech can mispronounce words or flatten important expression. For fluency, poetry, world languages, music, or content where pronunciation is part of the goal, provide a teacher-reviewed model.
Visual supports, concept maps, sequences, timelines, diagrams, and organizers can make relationships visible. They can also create barriers when labels are tiny, reading order is unclear, or the same idea is available only in the image.
For each meaningful visual:
Mark purely decorative images as decorative in formats that support that distinction. Do not ask AI to invent a scientific, mathematical, historical, or safety-critical diagram without checking every element.
A glossary, contextual definition, bilingual support, or translation can open access to the core idea. Review translations for subject meaning, tone, regional usage, names, and false cognates. Preserve key disciplinary terms when students are expected to learn them, and pair them with an explanation rather than replacing them everywhere.
Do not use home language as a public group label. Ask students and families about useful language supports through approved school practices, and treat language proficiency as developing, not as a measure of intelligence.
When available, the Annotation Suite may support highlights, notes, stickers, or voice notes on a resource. Use annotation for a defined cognitive job:
Teach one annotation convention at a time. Too many colors or stickers can increase visual and executive-function load. Confirm that an annotation can be created, found, edited, removed, and saved with keyboard or alternative input. Provide a text-based alternative when drag-and-drop or freehand placement is not usable.
Voice Access, dictation, audio responses, and microphone-based activities can support students who prefer or require speech input. AAC or symbol-supported options may be available in some learning tools.
Before using them:
Speech recognition errors are access failures, not student errors. Do not lower a grade because recognition software misheard an answer.
Whether the final format is a webpage, shared link, word-processing file, PDF, or printout:
A document that looks organized can still be structurally inaccessible. Visual inspection is only one part of review.
The HTML handout ships with its own accessibility toolbar, so many supports travel with the document instead of depending on the student's device: adjustable text size that reflows rather than clipping, offline dyslexia-friendly and high-legibility font options, reading themes, and annotations that stay anchored to the content through those changes. Printing always comes out black on white, whatever screen theme the teacher works in. Details and the font options table are in Documents and printing.
For students and staff using hands-free features, the microphone indicator names its state in text (Listening, Paused, Thinking, Speaking), not color alone, every change is announced to screen readers, and a small level meter confirms the microphone hears you. See AlloBot.
Use a table for data relationships, not page layout. Give it a descriptive introduction, identify header rows and columns in formats that support them, keep the structure simple, and avoid merged cells when possible. Check that the table is understandable when read one cell at a time.
For a complex table, provide a summary or another representation. On small screens, confirm that students can reach every cell without losing row or column context.
Alt text should communicate what the image contributes to the task, not list every visible detail. For example:
If students must analyze the image independently, do not put the answer in the alt text. Describe the data and relationships they need to perceive while preserving the reasoning task.
Provide accurate captions for speech and meaningful sounds, and a transcript when students may need a text route or search. Describe important visual information that is not evident from the audio. Avoid autoplay, provide playback controls, and check that keyboard focus does not become trapped in the player.
Review automated captions for names, vocabulary, punctuation, speaker changes, and timing. Generated captions are a draft.
Every interactive activity should have:
Do not use a game score, speed, streak, or animation as the only evidence of learning.
Check equations, units, superscripts, subscripts, chemical notation, charts, and diagrams in the final delivery format. Confirm that assistive technology announces them meaningfully and that a plain-text explanation or alternate notation is available when needed.
Do not replace formal notation when the notation is itself part of the goal. Teach students how to move between the formal representation and an accessible verbal description.
Keep the original source available when appropriate. Offer read-aloud, clarified layout, chunking, vocabulary, and a version that preserves essential content. If decoding or fluency is being assessed, separate access support for directions from the text students must read independently.
Provide a visible goal, numbered steps, estimated work segments, a model, a checklist, and a clear finish state. Use timers as optional planning aids, not public pressure. Let students resume work after interruption.
Reduce nonessential animation, sound, notifications, decorative movement, and crowded side panels. Offer a focus view or quieter route. Do not assume a visually playful tool is motivating for every student.
Support zoom and reflow, meaningful text alternatives, sufficient contrast, keyboard navigation, and screen-reader structure. Avoid directions such as “click the green item on the right.” Check the lesson without images to learn what meaning disappears.
Caption video, transcribe audio, and make alerts visible as well as audible. Give discussion questions in text. For live oral directions, provide a written version and face the class when speaking.
Avoid requiring precise dragging, rapid clicking, long handwriting, or speech as the only route. Provide selection, keyboard, switch-compatible, typed, partner-assisted, or teacher-recorded alternatives consistent with the student's plan.
Use models, visuals, bilingual resources, wait time, sentence frames, symbols, AAC, and multiple response modes as appropriate. Do not force public speaking as the only demonstration unless speaking is the target.
Set the mouse aside and:
Keyboard testing does not replace switch or screen-reader testing, but it catches many serious barriers.
Zoom the browser substantially and use a narrow window similar to a student device. Confirm text does not overlap, clip, or disappear and that controls remain reachable. Test the available themes, operating-system contrast settings when relevant, and reduced-motion preferences.
Do not judge contrast by eye alone for important materials. Use an approved checker or accessibility audit, then manually verify focus, states, charts, images, and meaning.
When the lesson will be used by a student who relies on a screen reader, speech input, magnification, switch access, AAC, captions, or another assistive technology, test with that technology and involve the student or appropriate specialist. A simulation or automated scanner cannot reproduce the student's full experience.
Do not change a student's established accommodations because another digital option appears convenient. Coordinate with the student's plan and support team.
Accessibility in the teacher workspace does not guarantee accessibility in an export, QR-linked assignment, embedded page, live activity, or printed copy. Test the final route:
Repeat the relevant checks after changes.
An Accessibility Lab, PDF accessibility workflow, contrast checker, or document audit can help find issues and suggest remediation. Automated results are valuable triage, but they cannot determine whether:
Review every automated fix. A change that improves a score can still damage content or usability.
Accessibility information can be sensitive. Do not paste an individualized education program, accommodation plan, evaluation, medical record, behavior note, or named student profile into a general generation prompt. Describe the design requirement without identifying the learner.
Offer supports in a way that does not announce who “needs” them. Normalize the use of tools, allow private choice or routing when possible, and avoid recording support choices as deficit data. Follow Privacy and responsible AI and local policy.
Ask students directly, in an accessible and nonjudgmental way:
Student feedback is evidence about the design, not a demand that the student justify an accommodation.
Goal and design:
Content:
Interaction:
Delivery:
For delivery planning, return to Prepare a lesson or continue to Live sessions. For reusable classroom patterns, see Classroom workflows.
AlloFlow can collect useful classroom signals, but a signal is not a verdict about a student. Use the dashboard, live-session summaries, quick checks, saved work, and your own observations together. The goal is to choose a helpful next instructional move while the context is still fresh.
If you have not yet run the activity, start with Live sessions. If a student may have encountered an access barrier, review Accessibility and UDL before interpreting the response.
Depending on the activity and deployment, you may see:
These sources answer different questions. An “opened” status is evidence of access, not understanding. A correct selected response may reflect understanding, a guess, or support from a peer. No recorded response may reflect absence, a connection problem, an inaccessible interaction, a different response mode, or a decision not to submit.
Write the target in student-observable language before looking at the dashboard. For example: “Students can support a claim with two relevant details,” not “Students completed the activity.”
This prevents completion data from quietly replacing the learning goal.
Ask:
Treat “no evidence” as a cue to investigate access first. Do not convert it automatically into a zero, a behavior judgment, or a mastery judgment.
Use Teacher Dashboard or the live summary to locate a possible class, group, or individual pattern. Then open several actual responses or work samples.
For a possible misconception, check whether students:
A dashboard pattern is a prompt for teacher review, not a diagnosis.
Before making a high-impact decision, use at least two kinds of evidence. Pair a selected-response result with a written explanation, conference, worked example, observation, or a second task in a different format.
If the sources disagree, gather a smaller, clearer sample instead of averaging the disagreement away.
Group students by the next support they need, not by fixed labels. Useful temporary groups include:
Change the groups when new evidence appears. A student can belong to different groups for different targets.
Prefer a focused response over regenerating an entire lesson. You might:
| Signal | It may suggest | Check before acting |
|---|---|---|
| Resource assigned but not opened | delivery or access problem | correct target, connection, device, session mode, and student presence |
| Opened but no recorded response | unclear prompt, access barrier, time, alternate response mode, or disengagement | observe, confer, and confirm where the response was captured |
| Many students choose the same distractor | shared misconception or ambiguous item | inspect the item and ask students to explain their reasoning |
| Strong selected responses but weak explanations | recognition without transfer, or an explanation barrier | use a worked example, oral conference, or alternate expression mode |
| Partial participation | time, load, navigation, language, or confidence issue | ask which step caused the stop and reduce the next task |
| Revision recorded | productive use of feedback | compare drafts for a meaningful change and name the successful strategy |
| One group performs differently | different support, pacing, prior knowledge, or task conditions | compare conditions before attributing the difference to students |
| No class-level pattern | mixed needs or insufficient evidence | inspect individual work and use a brief common follow-up |
Some deployments show an end-session preview before the teacher closes the room. Use it as a final instructional checkpoint.
Suggested cohorts are conveniences. Verify membership before sending anything, especially after students reconnect, change devices, or respond outside AlloFlow.
Teacher Dashboard may surface class summaries, trends, quality signals, concept patterns, or possible misconceptions. The available cards depend on which tools students used and on the deployment.
Start with a question:
Then trace the summary back to student work. Avoid copying a dashboard label into a permanent record without checking the evidence and following school policy.
When adding teacher comments, distinguish observation from interpretation:
This format is more useful than “low,” “unmotivated,” or “does not understand.”
At the end of the review, make three decisions:
Keep the follow-up proportional. A two-minute hinge question may be more informative than another full worksheet.
If you need the complete lesson state later, use Save Project to download an AlloFlow project file to an approved location. A project file may contain source text, generated resources, teacher notes, roster information, or student work depending on what you included. Give it a neutral filename and protect it like any other instructional record.
Export or download only what your school permits. Before sharing a dashboard, screenshot, CSV, PDF, annotation file, or project:
See Privacy and responsible AI for the full preflight.
Use this short structure in a planning document or approved record:
For ready-to-use follow-up ideas, continue to Classroom workflows.
These recipes turn common middle-school needs into short, repeatable AlloFlow workflows. They are starting points, not scripts. Keep the learning target constant while you vary access, practice, and expression.
Button names can differ slightly by deployment. If you cannot find a surface, press ? to turn on Help Mode, use Help Search, or return to the Launch Pad and choose Guided Mode, Learning Tools, or Educator Tools.
Before students use any generated material, complete the review steps in Prepare a lesson and the access checks in Accessibility and UDL.
Use when: Students need different entry points to the same article, passage, primary source, or teacher-created explanation.
Typical preparation: 10-20 minutes after the source is ready.
Students can enter through a version or support that reduces a barrier, but they still work toward the same essential understanding. Give students a route back to the original text for quotation, evidence, or productive challenge.
Ask, “Did I reduce unnecessary language load, or did I reduce the idea?” If the central concept or evidence changed, revise before sharing.
Use when: A student benefits from home-language access while building English vocabulary and participation.
Typical preparation: 10-15 minutes plus fluent review when available.
The student can use the home language to access meaning while seeing and practicing the English terms needed for class participation.
Do not treat machine-translated wording as proof of language proficiency or use translation quality as part of the student’s content grade. Confer when a response appears inconsistent across languages.
Use when: A middle-school learner needs targeted phonemic awareness, phonics, syllable, spelling, or handwriting practice without elementary-looking materials.
Typical preparation: 5-10 minutes.
The learner receives focused practice with immediate interaction and a visible connection to grade-level work.
Use performance to plan instruction, not to diagnose a reading disorder or replace a validated assessment. Coordinate sustained intervention with the reading specialist, special educator, or other responsible team member.
Use when: Students can discuss a text but need support selecting evidence, organizing a claim, or explaining reasoning.
Typical preparation: 15-25 minutes.
Students move from selecting evidence to explaining why it matters. They may type, dictate, record, draw and label, or use another approved expression route when that route measures the target.
Read the final work. Generated feedback can miss nuance, reward formulaic writing, or invent a source detail. Students should be able to explain their own claim and revision.
Use when: You need one clear decision before moving on.
Typical preparation: 5-10 minutes.
Students complete a short check whose purpose is to guide teaching. Tell them when it is ungraded and how you will use the result.
Do not grade a missing live response until you have ruled out delivery, access, language, device, and alternate-response issues. See Review evidence and plan next steps.
Use when: Students will rotate through simulations, models, data, engineering, coding, or other STEM experiences.
Typical preparation: 15-30 minutes, including a device test.
Students interact with a model or task while using a consistent reasoning routine. The technology supports investigation; it is not the learning target by itself.
Confirm that the simulation’s assumptions and scale are appropriate for the concept. A successful click path does not demonstrate scientific or mathematical reasoning unless students explain the result.
Use when: The instructional goal involves reflection, self-management, collaboration, conflict practice, digital well-being, or another school-approved SEL competency.
Typical preparation: 10-20 minutes plus review of the school response plan.
Students practice a bounded skill without being required to reveal personal or traumatic information. They know that an app is not a confidential counselor and that a trusted adult is the route for real help.
Follow the school’s crisis, bullying, mandated-reporting, and student-support procedures. Do not rely on automated safety checks as monitoring or emergency response. Review Privacy and responsible AI before using free-text coaching.
Use when: You are creating or adapting a handout, flyer, slide-like visual, PDF, audio, or video resource.
Typical preparation: 15-30 minutes, depending on the source.
Students can perceive, navigate, and understand the resource through more than one route without needing a separate “special” copy for routine access.
An automated audit can find some issues; it cannot certify that the resource is usable or compliant. Human review of meaning, reading order, captions, alt text, and interaction remains necessary.
Use when: The class shares a target but groups need different supports, examples, pacing, or response modes.
Typical preparation: 15-20 minutes plus a short live-session preflight.
Students receive the support intended for them without the teacher publicly announcing ability groups or sensitive accommodations.
Verify the recipient list before every targeted send. Group membership is an instructional convenience, not a permanent label. The complete run-of-show is in Live sessions.
Use when: A learner missed the live lesson or needs to finish after the session.
Typical preparation: 10-15 minutes.
The learner receives a coherent catch-up path rather than a stack of disconnected files. The work focuses on the essential target and includes a way to ask for help.
Do not assume that a generated link or code remains available after a live session ends. Test the independent delivery path before sending it.
Use when: A student communicates with symbols, partner-assisted scanning, a communication board, speech output, or another AAC system.
Typical preparation: Plan with the student’s communication team; simple activity adaptation may take 10-20 minutes.
The student can participate in the same thinking task through an established communication route and retain control over the message.
Coordinate durable boards, access methods, vocabulary changes, and goal tracking with the student, family, SLP, special educator, and other responsible team members. Generated symbols require human review for meaning, culture, age respect, and visual clarity.
After using a recipe:
For interpreting the resulting classroom signals, return to Review evidence and plan next steps. For safe storage, sharing, and AI use, continue to Privacy and responsible AI. Keep the Quick reference nearby while a routine is still new.
AlloFlow is designed to support data-minimizing classroom workflows, but the app name alone does not make a use private, compliant, or instructionally sound. The result depends on the deployment, provider settings, district agreements, content entered, sharing choices, and teacher practice.
Use this chapter as an operational checklist. It is not legal advice and does not replace district policy, an approved data-protection agreement, or review by privacy, security, special education, and legal staff.
Use the minimum information needed for the teaching task.
For routine generation, do not enter:
Replace a named student profile with an instructional description such as "a middle-school learner who benefits from shorter directions, visual examples, and a choice of typed or selected response." Remove rare combinations of details that could still identify the student.
Clinical or specialist tools do not create an exception to this rule. Only authorized staff should use identifiable records, and only in a deployment and workflow that the district has explicitly approved for that purpose.
If you cannot answer these questions, pause and use non-identifying content.
AlloFlow can run in more than one environment. Features and data handling differ.
| Workflow | What to assume until verified | Teacher action |
|---|---|---|
| AI generation | The prompt and source may be sent to the selected AI provider | Use de-identified content and confirm the provider is district-approved |
| Local or desktop generation | Processing may stay on the teacher device, but browser storage, project files, and local logs can still contain content | Use a managed device and approved storage; verify the actual backend status |
| Live Session | Session coordination, delivery, responses, or fallback data may use local peer connections or a configured backend | Use the district-approved live deployment and the approved codename or roster practice |
| Browser drafts and saved state | Work may persist on that browser or device | Do not use a shared profile for sensitive work; clear data according to policy |
| AlloFlow project file | The downloaded file can contain source material, generated resources, settings, notes, and other included state | Use a neutral filename and protect the file as an instructional record |
| Export, QR code, or share link | A new copy or access path is created outside the authoring view | Preview without teacher privileges, limit access, and remove the copy when no longer needed |
| Microphone, image, or portrait feature | Audio or images may be captured or sent to a provider depending on the feature and backend | Obtain required approval, provide a non-recording alternative, and verify the destination |
| LMS launch | Course, role, or assignment context may be supplied by the LMS | Use only the district-configured integration and follow LMS retention rules |
“Local” is not the same as “risk-free.” A lost laptop, shared Windows profile, synced Downloads folder, browser backup, or copied project file can expose local content.
The main live-session paths have different operational boundaries:
Do not infer that a feature is approved merely because it opens. Ask the district which URL, desktop package, AI backend, live-session mode, and storage location teachers should use.
For student sessions:
AI output is a draft. Before students see it:
Grounding or citation features can help locate sources; they do not eliminate the need to open the source and verify the claim.
Give an age-appropriate explanation:
Do not tell students that a free-text space is private, confidential, or continuously monitored unless the school has actually established and staffed that workflow. For sensitive SEL activities, state clearly that the tool is not a counselor or emergency service and point students to trusted adults and school procedures.
Do not use an AI generation, dashboard pattern, or automated score as the sole basis for:
Use validated instruments, required team processes, qualified professional judgment, direct evidence, and student or family participation as applicable.
Preview prompts, establish an opt-out, and avoid requiring personal disclosure. Automated safety checks can miss risk, misunderstand figurative language, or create false reassurance. Follow school crisis, bullying, mandated-reporting, and student-support procedures whenever a real concern appears.
Record observable context and behavior rather than intent or character. Separate facts from hypotheses. Behavior data can be highly identifying and sensitive; use the authorized team workflow and approved deployment. A generated pattern is not an FBA, BIP, diagnosis, or team decision.
Use prompts to observe strategy, response to support, and instructional next steps. Do not describe a generated activity as standardized, norm-referenced, diagnostic, or equivalent to an evaluation.
Only authorized professionals should use student records. Verify every score, descriptor, name, date, pronoun, source, and recommendation against the original record. Never let generated prose override test manuals, professional standards, team findings, or district document controls.
Generated symbols, boards, schedules, and stories require review by the student and communication team. Check cultural meaning, age respect, visual clarity, vocabulary placement, motor access, and voice ownership. Do not replace an established AAC system for convenience.
Automated checks and remediations are evidence, not certification. Confirm reading order, semantics, text alternatives, forms, captions, keyboard operation, and usability with human testing appropriate to the audience.
Before downloading or sharing:
An AlloFlow project is a convenient backup, but it is also a container of the content you put into it. Treat it accordingly.
If identifying or sensitive information is entered, sent to the wrong group, exposed in a share, or included in a bug report:
Do not paste the sensitive content into a support ticket. Use a redacted screenshot or a fictional reproduction of the technical problem.
Before generation or delivery, confirm:
For a practical live-room routine, return to Live sessions. For safe interpretation of results, use Review evidence and plan next steps. If a deployment or connection problem occurs, continue to Troubleshooting.
AlloBot is AlloFlow's built-in assistant. You can type to it or talk to it, ask questions about your lesson, or ask it to do things in the app. As of the August 2026 update, conversation comes first: you can speak to it naturally, and it will never scold you for not using a command.
If you are new to AlloFlow, read Start here first. AlloBot helps most once you have a lesson to talk about.
When hands-free mode is listening, everything you say lands in one of three buckets:
| What you said | What happens |
|---|---|
| Ordinary speech ("what should I add for my striving readers?") | AlloBot answers in conversation, out loud. There is no wrong thing to say. |
| Something that sounds like an action ("build a lesson", "open the glossary") | AlloBot offers first: it says what it could do and waits for your yes. Keep talking instead, and the offer quietly goes away. |
| A quick, harmless action ("bigger text", "stop reading") | It just does it. |
The offer step exists because a command that yanks you to a different screen mid-thought is worse than no command. Nothing that changes what is on screen runs without your yes.
The shortcut for people who know the commands: say "command" before a phrase ("command open the learning hub") and it acts immediately, no offer.
The voice indicator always tells you, in text as well as color, which state it is in: Listening, Paused, Thinking, or Speaking. A small level meter shows your own voice being picked up, so you are never left wondering whether the microphone hears you. All of it is announced to screen readers.
The X on the bot and the header toggle do the same thing: they hide AlloBot and its tips. Nothing else. Read-aloud, narration, and any speech you ask for keep working with the bot hidden.
A spoken question goes to the same AI backend as everything else you generate, under the same settings. AlloBot does not listen when hands-free mode is off, and the microphone closes while it speaks or thinks (the indicator says so). If no AI backend is set up, AlloBot's AI conversation is unavailable; see Troubleshooting for the three ways to turn AI on.
Everything AlloFlow generates can leave the app as a document a student can hold: a finished copy to read, a blank copy to write on, or a self-contained web page with its own reading tools. This chapter covers choosing between them and the details that matter once paper is involved.
For building the resources themselves, see Prepare a lesson.
| Format | What it is | Reach for it when |
|---|---|---|
| Print / Save as PDF | A finished copy to read. Opens your print window; print it or save as PDF there. | Reference copies, family communication, reading material. |
| Worksheet | A blank copy to write on: answer boxes become ruled lines, multiple choice becomes fill-in bubbles, and a Name / Date / Score header appears. | Paper practice and anything handed in physically. |
| HTML handout | One self-contained file with its own reading toolbar. Works offline once downloaded; nothing phones home. | Sending work home, devices without AlloFlow, screen-reader use. |
| Slides | A presentation export. | Whole-class instruction. |
The export screen explains each format under its selector, so you do not need to memorize this table.
The worksheet has one more trick: if your lesson has both an adapted text and a glossary, you can turn the passage into a fill-in-the-blank worksheet. Glossary terms become numbered blanks, a word bank appears below, and the teacher copy carries the answer key. It works in any lesson language.
The exported page carries its own accessibility toolbar, independent of AlloFlow:
Students and teachers can highlight, add notes and stickers, record voice notes, and draw freehand directly on the handout. Marks belong to the document: they scroll with it, stay attached through text-size and font changes, and survive reload. "Erase my marks" clears only your own annotations, tells you how many it would erase, and is disabled when there is nothing to erase.
Teacher annotations made in the app travel into the export, including drawings.
The Document Builder's ribbon has an Expert Workbench tab: type a plain-language instruction ("make every heading a proper H2", "add alt text to the images") and it applies structural edits across the document. It is the same engine the PDF remediation pathway uses. The panel shows example instructions, so an empty text box never greets you.
Schools run on inherited documents: the scanned packet, the district PDF, the worksheet someone made in 2009. AlloFlow's remediation workflow turns a document like that into an accessible version, shows its evidence, and keeps you, the human, as the last step. This chapter is the teacher's view; the white paper covers the same pipeline for a district evaluation.
A thorough run on a long document takes a while, and there are three things worth knowing.
If you already use Claude Desktop, there is a local connector that runs the same pipeline on your own computer. You install it once, and then you can ask in plain language: audit this document, remediate it, export it as EPUB. It reads the file from your disk, so nothing is uploaded to AlloFlow.
Two things make it worth knowing about. The deterministic tools (validation, text extraction, redaction, structure checks, exports) work with no AI key at all, and the AI-assisted repair runs on a key you supply yourself. Your IT department may prefer this path for exactly that reason: see For your IT department.
Every run produces a report bound to the exact files it describes: what was found, what was fixed, what remains, and which checkers said so at which versions. Keep it with the document. If anyone ever asks "how do you know this version is accessible," the answer is a report, not a recollection.
The whole workflow can run against a local AI model instead of a cloud provider, and the validation tools run locally, so a sensitive document can be processed with nothing leaving the machine. Set that up in AI Backend Settings ("Private AI on this computer"), and see Privacy and responsible AI for the handling rules that still apply to the files themselves.
When something breaks during class, protect student learning first and diagnose second. Move students to the prepared fallback, preserve unsaved work, and avoid turning a connection problem into a participation or behavior judgment.
If the issue involves identifying or sensitive information, stop sharing and follow the incident steps in Privacy and responsible AI.
Use the safest available option:
If Save Project is unavailable, keep the tab open while you test a second approved browser tab or device. Do not copy student data into a public chatbot, personal email, or unapproved note service.
The right fix depends on where AlloFlow is running.
| Environment | First checks | Escalate when |
|---|---|---|
| District-hosted web app | approved URL, sign-in, online indicator, browser support, district service notice | several users or features fail on the approved site |
| Desktop or local-network app | desktop runtime is open, local engine status, teacher host is reachable, devices are on the intended network | the runtime will not start or school security controls block the local host |
| Canvas-style or provider-hosted authoring | school account, platform availability, correct shared artifact, provider limits | the surrounding platform fails or student live use is not part of the approved design |
| LMS launch | course and role context, pop-up or new-tab behavior, third-party cookie restrictions, assignment link | the launch has the wrong role/course or repeatedly loses context |
| Student live-session page | correct environment, active code, approved codename, connection status | multiple students cannot join, receive, or return responses |
Use one path at a time. Do not mix a browser-hosted URL, a Desktop LAN address, and a School Box address in the same class session.
After identifying the path, repeat the same controlled test: one teacher action, one fictional or low-risk resource, and one student/device. If that test fails, stop changing variables and escalate. Do not change firewall, proxy, certificate, device-management, or account settings on your own. Give district IT the deployment and symptom details.
If a content filter or network rule is suspected, report the approved site URL, the visible tool name, the time, and the error. IT can inspect the request without receiving student content.
That small sparkle pill means no AI backend is set up on this device: no API key, no local model, and you are not inside Gemini Canvas. Nothing is broken. Every sim and activity works fully without AI; only the extras (AI hints, coaching) are off. Click the pill to see the three ways to turn AI on. The easiest is opening AlloFlow inside Gemini Canvas, which is free with a Google account and uses your Gemini plan's daily quota; the setup screen has a button that takes you straight there.
Notices appear at the top center of the screen and fade after a few seconds. Nothing is lost: click the lightbulb in the header and open the Messages list to reread every recent notice, newest first.
Possible causes include provider availability, account or quota restrictions, local-engine status, a network interruption, an oversized or unsupported source, or a deployment configuration problem.
Do not solve the problem by repeatedly requesting “make it better.” Tighten the task:
Then perform the human review in Privacy and responsible AI. A fluent answer can still be inaccurate.
Do not upload secure tests or copyrighted material that the school is not permitted to process or redistribute.
Check in this order:
If several students cannot join, stop cycling codes during instruction. Use the fallback and test a new session after class.
Do not publicly identify a support group while troubleshooting. Verify group membership privately.
Missing data is not proof that a student did not work.
Use Review evidence and plan next steps before interpreting an incomplete Activity Pulse or end-session summary.
Microphone permission does not guarantee that a recording or speech service is available. Local browser speech support, the selected voice provider, and district network policy can differ by device.
Never require a student to record a face or voice when the workflow, consent, or storage path is not approved.
A project may contain sensitive instructional state. Do not attach it to a general support ticket. Provide a de-identified reproduction when possible.
If a public or overly broad share was created, revoke it and follow the privacy incident procedure.
Run Accessibility Lab when available, but also test the real student path. Report the exact control label and the keyboard step where the barrier occurs.
Educator Tools such as BehaviorLens, Report Writer, Symbol Studio, Accessibility Lab, document pipelines, and media studios may load additional components.
Include:
Do not include names, session rosters, live codes, student responses, project files, assessment records, disability information, API keys, or unredacted screenshots.
Move fully to the fallback and contact the appropriate school support when:
After class, reproduce the issue with fictional content and one controlled variable at a time. For routes and terminology that can help you describe the problem, see Specialist and product reference.
AlloFlow includes a math fluency instrument: short, timed practice in the curriculum-based measurement (CBM) style, plus a Fluency Maze game. It records attempts and gives you a score history, which makes it closer to an assessment tool than to the exploratory STEAM Lab sims.
Math Fluency is a mode inside the Math tool. Open the Math card in the sidebar, then change the Mode selector to Fluency Probes or Fluency Maze.
The faster door, added August 2026: press Ctrl+K and type any of these: "math fluency", "fluency probe", "timed math", "math minute", "CBM probe", or "math maze". The command palette takes you straight there. Because opening a timed probe changes what is on screen, the assistant confirms before it acts.
The panel lives in the teacher sidebar, so students do not browse to it on their own. You administer a probe, or you deliver it as part of an assignment. Teachers, independent learners, and family mode can all reach it; the student view cannot.
For the assessment side of the house, including probe administration from Assessment Center, see Specialist and product reference.
AlloFlow includes a core lesson workflow and several specialist, accessibility, clinical-adjacent, creative, and technical surfaces. This chapter helps you choose an entry route and understand the boundaries. It is intentionally organized by purpose rather than by a fixed tool count, because the catalog and labels can change.
For a first lesson, begin with Start here. For a practical teaching need, use Classroom workflows.
| Route | Best starting point for | What to expect |
|---|---|---|
| Guided Mode | planning and delivering one differentiated lesson | a structured sequence from source through directions, resources, review, and delivery |
| Learning Tools | opening a focused student-facing literacy, STEM, SEL, study, or creative experience | hub and tool choices without building a complete lesson first |
| Educator Tools | teacher, specialist, document, accessibility, behavior, AAC, reporting, and professional workflows | professional surfaces that may require additional policy, training, or configuration |
| Full AlloFlow (sometimes called Full Platform) | experienced users who know the destination | the broad navigation and complete set of available surfaces |
If the app opens directly into another view, use the header or home control to return to the Launch Pad. Labels may vary slightly by deployment.
Do not rely on an old screenshot to locate a feature. Hubs and routes evolve faster than the teaching workflow.
| Teaching or professional need | Start with | Use with care |
|---|---|---|
| set the grade, language, and other defaults for a whole lesson | Universal Settings | it applies to new work only, so resources you already made keep their old settings |
| adapt a source, build directions, or create a lesson package | Guided Mode, Source Material, configuration, lesson resources | verify every generated claim and preserve the common learning target |
| create a picture or diagram for the content | Lesson Images | check depictions for accuracy and stereotype, and add a text alternative |
| read, understand, or discuss text | Text Adaptation, Glossary, Visual Organizer, Immersive Reader, annotation tools | do not confuse a simpler text with a simpler learning goal |
| check understanding | Quiz, Exit Ticket, Polling, Concept Sort, live activities | use as formative evidence unless a reviewed assessment plan says otherwise |
| teach a synchronized class | Live Session, Group Manager, Teacher Signal controls, Activity Pulse | verify recipients and treat missing responses as possible access problems |
| review class patterns | Teacher Dashboard, learner progress, session summaries | trace summaries back to actual work before making decisions |
| practice sounds, spelling, or word patterns | Word Sounds Studio | use age-respectful materials and coordinate intensive intervention with specialists |
| explore mathematics, science, engineering, coding, or data | STEAM Lab and saved STEAM stations | preview simulations and require reasoning, not just successful interaction |
| teach reflection, relationships, self-management, or digital well-being | SEL Hub and saved SEL stations | do not require personal disclosure; follow school safety procedures |
| support AAC, visual communication, schedules, or social narratives | Symbol Studio | preserve the student’s established system and involve the communication team |
| collect and review observable behavior data | BehaviorLens | authorized team use only; no automatic diagnosis or plan |
| observe response to prompts and scaffolds | Dynamic Assessment | not standardized or diagnostic unless a separate validated process establishes that |
| draft or organize formal report content | Report Writer | authorized professionals must verify every fact, score, interpretation, and recommendation |
| inspect learner access | Accessibility Lab and student preview | automated checks support, but do not replace, human usability testing |
| inspect or remediate a document | PDF Accessibility and document pipeline tools | verify the final file independently; an audit is not certification |
| create accessible classroom graphics or worksheets | AlloStudio | keep real text, reading order, alt text, contrast, and source rights intact |
| create video, cinematic, story, performance, poetry, or music experiences | Video or Cinematic Studio, StoryForge, LitLab, PoetTree, Open Groove Studio | review age appropriateness, captions, flashing or motion, copyright, consent, and attribution |
| research, study, or prepare for a test | Research Hub or lanes, Learning Commons, Test Prep surfaces | verify sources and avoid secure test content |
| configure a deployment or school workflow | Admin or IT documentation and deployment controls | restricted to authorized staff; test privacy, security, retention, and accessibility settings |
BehaviorLens can support structured observation, data collection, visualization, and team discussion. Use objective descriptions of what occurred, the context, and the measured dimension. Avoid labels such as “defiant,” “manipulative,” or “attention seeking” when the record does not establish function.
It does not turn a single observation, generated hypothesis, graph, or preference activity into a completed FBA, BIP, diagnosis, or placement decision. Follow district consent, team, qualification, retention, and record procedures.
Dynamic Assessment surfaces can help an educator observe how a learner responds to prompts, models, feedback, or graduated support. Record the task, support provided, response, and change.
Do not compare generated results with standardized norms unless the actual instrument and manual authorize that interpretation. Use the information to form instructional hypotheses and decide what to observe next.
Report Writer can help organize authorized evidence and draft prose. The responsible professional remains the author.
Before a report leaves draft status:
Do not place identifiable records into an unapproved AI backend or deployment.
Symbol Studio can support communication boards, visual schedules, social narratives, symbol-based activities, and other visual supports. Generated content must be reviewed for meaning, culture, age respect, motor access, visual complexity, and consistency.
The student’s established vocabulary organization, access method, and voice take priority. Durable changes should be made with the student and the responsible AAC or communication team, including the SLP, family, and special educator as applicable.
Accessibility Lab helps teachers preview barriers and inspect the learner experience. PDF and document tools can identify and repair some structural, visual, or text-alternative issues.
No automated result proves that a resource works for every learner or satisfies a legal standard. Test the final student path: keyboard, focus order, headings, reading order, form labels, text alternatives, captions, zoom, reflow, contrast, timing, and comprehension.
SEL Hub supports instruction and practice; it is not confidential therapy, crisis monitoring, or a substitute for a school mental-health professional. Preview prompts, define privacy expectations, provide an alternate route, and avoid forced disclosure.
If a student communicates a real safety concern, stop treating the exchange as an app activity and follow the school’s human response procedure.
AlloStudio, video and cinematic tools, StoryForge, LitLab, PoetTree, and Open Groove Studio support student creation and multiple forms of expression.
Before publication or performance:
Research tools can help organize questions, sources, notes, and synthesis. Open every cited source and check author, date, evidence, context, and relevance. Generated citations can be incomplete or incorrect.
Test-preparation tools should teach concepts and strategies. Do not upload secure items, reproduce restricted forms, or describe a generated practice score as an official result.
| Item | What it is | Teacher implication |
|---|---|---|
| Source Material | the text, file, URL content, or prompt used to ground the lesson | use permitted, de-identified content and keep the authoritative original |
| Generated resource | a draft lesson item such as adapted text, glossary, quiz, organizer, or activity | review before students see it |
| History | the resource list or activity state in the current project | titles and old drafts may still expose sensitive content |
| Browser draft or saved state | local or synced state used to resume work | shared devices and profiles need an approved clearing routine |
| Saved resource pack | a generated pack kept on this device, reached through Open saved work | restore, pin, export, or erase it there; on a shared device, erase before handing it on |
| AlloFlow project | a downloadable project file containing included lesson state | store and share it like the most sensitive content inside it |
| Live Session | a temporary classroom connection for delivery, pacing, activities, and evidence | use the approved deployment, recipients, code practice, and close-out routine |
| Session summary | a record or local summary of participation and instructional signals | verify completeness and follow retention policy |
| Export or share | a new student-facing file, package, QR path, or link | preview access and permissions outside teacher mode |
| Annotation file | saved student or teacher notes, highlights, or voice-note references | review for identifying or sensitive information before exchange |
| Dashboard | an aggregate or per-learner view derived from available activity data | use patterns to ask questions, not to create permanent labels |
Save Project is the dependable teacher-controlled backup when available, but it is not a substitute for a district record system. See Privacy and responsible AI.
| Action | Common control |
|---|---|
| Move forward through interactive controls | Tab |
| Move backward | Shift+Tab |
| Activate a focused button or link | Enter; Space for many button-like controls |
| Close the current modal or overlay | Escape when supported |
| Toggle contextual Help Mode | ? |
| Open Help Search | Ctrl+K where enabled |
| Move within a supported menu, tour, slider, or grid | Arrow keys as announced by the control |
| Silence app audio | Mute All Audio in the header |
| Adjust reading presentation | reading theme, text, spacing, ruler, highlighting, or Immersive Reader controls as available |
Keyboard behavior can vary inside simulations and specialized editors. Turn on Help Mode, read the on-screen instructions, and provide an alternate route if a control cannot be operated with the student’s access method.
AAC: Augmentative and alternative communication. Methods and systems that support or replace speech, including symbols, text, gestures, partner-assisted scanning, and speech-generating devices.
Activity Pulse: A teacher-facing live view of activity lifecycle or response status. It helps identify delivery and follow-up needs; it is not a mastery score.
Adapted Text: A version of the source rewritten toward a target reading level. Produced by the Text Adaptation tool. It should change the reading load, not the learning goal.
AI backend: The configured service or local engine that performs generation. Provider, account, data handling, availability, and capabilities depend on deployment.
AlloBot: The in-app assistant. It answers questions in conversation and can offer to open a tool for you. Screen-changing actions are offered rather than performed, so you confirm before the view moves.
AlloFlow project: A downloadable project file used to preserve lesson state for later loading. It may contain source and generated content and should be stored accordingly.
Assignment or package: A reviewed collection of directions, resources, and evidence tasks delivered through an approved share, export, LMS, QR, or live route.
Codename: A classroom alias used to reduce the need for student names in a live workflow. A codename still requires careful handling if the teacher can link it to a student.
Deep link: A direct web address that opens one interactive tool by itself, with no sign-in (for example, a water cycle link a teacher can put in any assignment or message). Every STEAM Lab tool has one.
Deployment: The approved combination of app location, hosting, accounts, AI backend, live-session services, storage, security rules, and retention settings.
Expert Workbench: A Document Builder ribbon tab that applies plain-language structural instructions to a document ("make every heading a proper H2"). The same engine the PDF remediation pathway uses.
Evidence cohort: A temporary grouping suggested from recorded classroom signals, such as incomplete participation, catch-up need, or revision. The teacher must verify the group and its next move.
Generated resource: AI-assisted or tool-produced material that requires teacher review before use.
Guided Mode: A task sequence that leads from source and lesson intent through resources, directions, review, and delivery.
Help Mode: Contextual explanations attached to supported controls, toggled with ?.
Immersive Reader: A reading overlay that can provide read-aloud, highlighting, spacing, focus, or other presentation supports depending on the current build.
Lesson Images: The image generation tool. It creates pictures and diagrams for the content, and can refine or remove text inside an image. Do not confuse it with Visual Organizer, which builds concept maps and diagrams of structure, or with the AAC visual supports used for communication boards and schedules.
OCR: Optical character recognition, which converts an image of text into machine-readable text. OCR output must be checked, especially for tables, formulas, columns, and poor scans.
Source Material: The grounding content for generation. It remains the reference for fact and quotation checks.
Station: A saved or assigned learning-tool experience used in a rotation, center, or independent sequence.
Student-Paced: A live-session mode in which students can move through assigned material with more individual control.
Teacher-Paced: A live-session mode in which the teacher advances or activates the shared classroom experience.
Teacher Signal: A live-session cue, direction, delivery, or status control used to coordinate the class. Confirm the audience before sending.
TTS: Text-to-speech. Voice and language availability depend on the browser, provider, device, and deployment.
UDL: Universal Design for Learning, a planning framework that anticipates variability through multiple routes for engagement, representation, and action or expression.
Universal Settings: The shared defaults every generator inherits, including grade level, output language, translations, standards, interests, Depth of Knowledge, and emoji. They apply to new work only. See Universal Settings.
Worksheet export: The export format that produces a blank copy to write on: ruled answer lines, fill-in bubbles, and a Name and Date header. Contrast with Print / Save as PDF, which produces a finished copy to read. See Documents and printing.
For recovery steps, see Troubleshooting. For safe interpretation of dashboard and live evidence, see Review evidence and plan next steps.
AlloFlow is easiest to sustain when a school adopts a small number of repeatable instructional routines rather than asking teachers to learn every tool. This chapter is for principals, instructional coaches, department leads, special educators, and teacher teams. It can be adapted into a grade-level launch plan or a short professional-learning sequence.
Before choosing a feature, agree on what teachers and students should experience:
Write the promise in ordinary staff language. For example:
We will use AlloFlow to make worthwhile work more reachable, give teachers better choices for differentiation, and preserve teacher judgment about content, evidence, and student support.
Do not begin with a target such as “every teacher will use ten tools.” Feature counts are not evidence of instructional improvement.
A pilot answers one question: does this help our teachers with our students? Three to five volunteer teachers for four to six weeks is enough to know.
Keep the entry cost near zero. Every interactive tool has its own direct link (for example, a water cycle simulation a science teacher can open with one click and use that period). Start volunteers with one link that fits something they already teach; nobody has to adopt a platform to try a tool. There are no accounts to create, for staff or students, so IT setup for the pilot itself is nothing.
Agree on defaults once. Have the pilot group open Universal Settings together for ten minutes and settle grade level, language, and the translation setting for their context. Shared defaults make pilot artifacts comparable and prevent the most common early confusion, which is two teachers getting different-feeling output from the same tool.
Give each role its chapter. Volunteers get Start here and Prepare a lesson. Whoever supports them gets Troubleshooting. The person leading the pilot reads this chapter and Privacy and responsible AI before the first classroom use, because privacy questions arrive on day one and deserve a prepared answer.
Close the loop weekly. A ten-minute standing check-in beats a survey: what did you try, what held up, what got in the way. Route "it broke" items to whoever maintains your deployment and "it confused me" items into the next check-in agenda. Both kinds are pilot data.
Decide in advance what would count as success, and keep it instructional: a teacher reuses a tool without being asked, a student who avoided a task engages with an adapted version, a co-teacher borrows a pilot teacher's material. The measures section later in this chapter has more, but three concrete stories beat a dashboard at this scale.
Leadership and the implementation team should:
The success condition for the first week is not polished production. It is that a teacher can explain the goal, source, review gate, student route, and fallback.
In a planning meeting:
An implementation team should model the actual learner route, not only demonstrate the teacher workspace.
Use Live sessions to rehearse a short routine:
Use fictional codenames and a small test group. Rehearse the standard live path and the school-approved alternate path. Decide who helps when a student cannot join, a resource does not arrive, or a connection fails.
Review the live-session protocol on GitHub with the staff member responsible for deployment support. Never make a live lesson depend on an untested network path.
Ask teachers to bring one example of:
Use Review evidence and plan next steps to keep the conversation about instruction. Do not use a dashboard card, participation count, or generated quality signal as a stand-alone evaluation of a student, teacher, or program.
Keep the meeting short and concrete:
Require teachers to bring a student work sample or an anonymized description, not a screen full of private student data. A useful PLC leaves with one lesson change, not a list of features to explore.
An instructional coach can look for these seven indicators during a planning conference or lesson:
| Indicator | Observable evidence |
|---|---|
| Purpose | The learning goal and success evidence are visible. |
| Source fidelity | Generated or adapted material preserves important facts, qualifiers, examples, and citations. |
| Access | At least one barrier has been anticipated and a usable support or alternate route is ready. |
| Agency | Students can make a meaningful choice without being publicly sorted or labeled. |
| Review | The teacher has checked the final student-facing route. |
| Privacy | Content, codenames, sharing, and storage follow the school’s approved boundary. |
| Recovery | A human-readable fallback exists if the provider, connection, device, or module fails. |
Use the rubric for coaching questions, not compliance scoring. A teacher may deliberately use a simpler route because it best fits the lesson.
The teacher owns the learning goal, source selection, generated-content review, student directions, accessibility check, live pacing, and interpretation of evidence.
The coach helps plan barriers, response options, accessibility checks, and follow-up questions. Specialists should not be expected to repair every technical issue during instruction.
The deployment lead confirms approved URLs or desktop builds, accounts and permissions, AI-provider configuration, network requirements, browser or device support, backups, and the escalation route. Start with the deployment guide on GitHub when the school owns or manages its host.
The principal or program lead protects planning time, sets the privacy boundary, identifies a small pilot, listens for friction, and decides which routines deserve broader support. Leadership should not request identifiable student exports merely to prove adoption.
Useful early signals include:
Interpret cautiously:
Pair any quantitative signal with teacher explanation and student work. Keep evaluation of personnel and formal assessment outside automated summaries.
Every teacher should be able to answer these questions without opening a feature catalog:
Keep the answers on a single page and write them for this school, not for AlloFlow in general. Questions 1 through 5 belong to the teacher and change with each lesson. Questions 6 and 7 are school decisions and should be filled in once by leadership, in specific terms: the approved deployment and student link, the storage location, the named support contact, the escalation route, and the boundary on what may be entered or shared.
Post the page where planning actually happens and revisit it whenever the deployment, provider, accounts, or approved routes change. A teacher who cannot answer question 6 or 7 has been handed a gap in the rollout, not a personal deficiency.
Treat the page as an orientation aid rather than a compliance form. It exists so a teacher joining mid-year, a substitute, or a specialist supporting a single student can act confidently without reading this guide first. If answering all seven questions takes more than a few minutes, the routine is probably too elaborate to sustain, and the more useful response is to narrow the routine rather than expand the page.
When a school pilots AlloFlow, the questions come from IT within a week. This chapter answers them in IT's own terms. It describes the public deployment; if your district self-hosts, verify each answer against your own instance, in the spirit of Privacy and responsible AI: assume nothing until verified.
A static web application served from a CDN. There is no application server holding user data, no database of students, and no login system. The code is open source (AGPL) and publicly auditable on GitHub, so your security review can read exactly what runs.
None. Staff and students use it without creating accounts. This removes the usual pilot blockers (rostering, SSO integration, account lifecycle) and also means AlloFlow holds no credential or directory data at all.
Work product stays in the browser's own storage on the device that created it (localStorage and equivalent browser storage). Nothing is synced to an AlloFlow server, because there is not one. The practical consequences cut both ways:
Browser storage is the default, not the only destination. These are the paths an audit should account for, in rough order of how often they are used:
| Destination | Who owns the storage | What lands there |
|---|---|---|
| Browser storage on the device | The device, under your fleet policy | All working state: workspaces, drafts, settings, generated resources |
AlloFlow project file (Save Project) |
Wherever the user saves it | A full snapshot: source material, generated resources, settings, notes |
| Exports (PDF, print, accessible document, worksheet, teacher copy) | Wherever the user saves it | Finished copies, and the teacher copy also carries answer keys |
| QR code or student share link | No new storage; a new access path | Whatever the shared route exposes |
| Live session transport | Local peer connections, or a mailbox you deploy (below) | Session coordination and student responses |
| LMS launch | Your LMS | Course, role, and assignment context supplied by the LMS |
| Optional Apps Script services | Your school's own Google Drive | Only what that specific service is for (below) |
| On-device model and remediation caches | The device | Cached speech models and remediation working data |
Two consequences worth stating to a leadership team. Browser storage means a wiped device or a cleared profile destroys unexported work, so "save the project file" is an operational instruction, not a nicety. And "local" is not the same as "safe": a shared Windows profile, a synced Downloads folder, or a copied project file moves instructional content exactly as far as any other file would.
The full teacher-facing data-path table is in Privacy and responsible AI; it is written to be reviewed line by line with a privacy officer. The teacher-side mechanics of saving, exporting, storage presets and recovery are in Saving, loading, and managing storage.
AlloFlow has no server. When a school genuinely needs shared or persistent storage, the pattern is always the same: a small Apps Script project deployed into a school-owned Google account, never a personal one, so the data sits in your Drive under the Workspace for Education agreement you already hold. Each is optional, and each is off until someone deploys it. The packages and their deployment READMEs are in apps_script/.
They do not share one security model, and the differences are the part worth your attention:
| Service | Deployed by | Identity model | Holds |
|---|---|---|---|
| Educator Evaluation repository | District administrator | Google identity, locked to your domain; the server decides each person's role | Personnel records, with a tamper-evident audit chain |
| Walkthrough Records | A principal, in their own account | drive.file scope only; each file is Restricted and shared with one named teacher, so Google enforces the reader's identity |
Walkthrough feedback a human wrote and approved |
| Class Mailbox | A teacher, in their own account | Capability tokens, no student accounts; possession of the link stands in for identity | Live session and homework-pack traffic |
| Leadership Hub backup | A school leader, in their own account | Capability token that can only touch files the script itself created; files created Restricted | The Leadership Hub's own backup file |
The distinction that matters: a link-possession model is appropriate for anonymous class traffic and inappropriate for anything about a named staff member. The Class Mailbox is deliberately built the first way and the other three are not. Walkthrough Records is explicitly not a system of record and never scores anyone; if you need a district system of record with verified identity, assignments and an audit trail, that is the Educator Evaluation repository.
The one service that holds personnel records deserves a named pointer rather than a summary here. It is optional; without it the evaluation tool keeps records in the signed-in browser profile on one device like every other tool. When a district does deploy it, an administrator installs it into a district-owned Google account, never a personal one, and from there it runs in your tenant, makes no external network calls, fails closed to accounts on your domain, sets its storage private and verifies that it took effect, and sends notifications that carry no ratings or evidence. The legal frame is personnel law, your collective bargaining agreement and district retention policy, not FERPA. You can verify a deployment yourself with verifyDeploymentIdentity(), getPortalSetupHealth(), verifyAuditChain() and doGet?api=health.
For the full treatment, read the Educator Growth & Evaluation user manual, which covers setup, roles, the audit chain and the release workflow in detail, and For school leaders for how it sits beside the rest of the suite.
AI features only work when a backend is configured, and the district controls which:
| Option | What leaves the device | Fits when |
|---|---|---|
| Gemini Canvas or a Gemini API key | Prompts and source content go to Google under the account or key used | The district already permits Google AI services |
| A local model (LM Studio, Ollama, LocalAI on the device or a district server) | Nothing leaves the machine or the district network | Confidential contexts, strict no-egress requirements |
| No backend | Nothing; AI features hide or disable themselves and everything else keeps working | Evaluation phase, or a deliberate non-AI deployment |
Teachers are instructed throughout this guide to use de-identified content with any cloud backend. The no-backend state is honest by design: the interactive tools, delivery, and documents all work without AI.
alloflow-cdn.pages.dev).Code.gs for UrlFetchApp before you approve it. There are zero occurrences, which is the claim that nothing leaves your tenant, and it takes one search to confirm rather than trust.AlloFlow is a static, open source web app with no accounts and no vendor-side data storage; work stays on the device by default, and AI runs only through the backend the district chooses, including a fully local option. Where a school does want shared storage, it deploys a small optional service into its own Google Drive, so that data stays in your tenant too. The realistic review effort is the same as approving a website plus, if you choose one, an AI provider you have likely already reviewed.
Principals, coaches, and student-services leaders get their own tool suite: open Educator Tools and choose the Leadership Hub 🏛️ card. Nine tools live behind it. This chapter explains what each one is for, how a first session goes, and the boundaries each tool deliberately keeps — because in this suite the boundaries are the product.
If you have not read Privacy and responsible AI, read it first. Everything below assumes its ground rules.
The hub states this once, and every tool holds to it:
A practical consequence worth planning around: because the data is on-device, a leadership workflow lives on the device where you started it. The hub's Back up this hub section covers this two ways:
apps_script/leadership_hub_backup/README.md in the AlloFlow repository; the script can only write files it created — it cannot read the rest of your Drive and never shares anything with anyone.Either way, the backup is working continuity, not filing. When a record needs to reach wherever your district officially stores it, download it from Drive and share it through the normal district channel yourself — that handoff stays deliberate and human.
Growth-framed classroom visits scored against UDL 3.0 look-fors, with feedback cards for teachers, a building heatmap, trends over time, and inter-rater checks.
First session. Open Roster & setup and add your classrooms — a teacher name is optional, and either way each classroom gets a code that the analytics use. Then open Observe, pick the classroom, and record what you see against the look-fors during a ten-minute visit. The visit lands in Visits, and Building starts filling in once you have a handful.
Keep in mind. The heatmap describes patterns in your observations, not teacher quality; two observers can calibrate with the inter-rater view before you trust a trend. Share feedback cards as conversation starters, not scores.
Turns your shorthand walkthrough notes into evidence-based coaching feedback — that you edit and approve. Every claim it drafts is cited back to a line in your own notes; anything it cannot anchor to your notes is flagged rather than trusted. It also includes practice scenarios for building the note-taking habit.
First session. Try a practice scenario before a real visit: it hands you sample notes so you can see how drafting, citation, and your approval step work without any real classroom involved.
Keep in mind. Formative coaching only. The Copilot never assigns a rating, never computes an evaluation score, and never makes an employment recommendation — by design, not by omission. If a draft sentence has no citation to your notes, that is the tool telling you it invented something; delete or rewrite it.
The card behaves differently depending on your district. If your district runs a connected evaluation portal, the card opens it — sign in there, and records live there. Without a portal it opens a private on-device workspace with completion and weighting views, walkthroughs, formal observations, SPM/SLO tracking, dialogue, receipts, and audit history, with framework profiles for PA Act 13, Portland (Maine) PEPG, and a general Maine PEPG profile that mirrors your district plan.
Keep in mind. The on-device workspace holds real working records — it is not a demo. What you enter stays in your signed-in profile on that device and is never uploaded, which on a managed 1:1 fleet means it is scoped to you and encrypted at rest by the device itself. AlloFlow adds no encryption of its own, so device sign-in and disk encryption are what protect it; that is a question for your technology director, in For your IT department.
Two things the on-device workspace genuinely cannot do, which is what the district portal adds: educators cannot sign in to see and acknowledge their own records, and there is no shared, retained, discoverable store for the district. Those are records-management needs, not a reason to keep real work out of the tool. Apply your district's retention rules to what you keep here, and back it up.
A note on the legal frame, since it is easy to reach for the wrong one: evaluation records are personnel records, so what governs is your state's personnel-records law, your collective bargaining agreement, and district retention policy — not FERPA, which covers student education records. FERPA enters only by the side door, if observation evidence names identifiable students, which is a good reason to keep written evidence focused on educator practice.
Evaluation is the one tool in this suite complex enough to have its own dedicated manual, and this chapter deliberately does not duplicate it. For the full walkthrough — choosing between workspace and portal, the evaluation cycle step by step, framework profiles, portal deployment, how released evaluations reach teachers, and its own troubleshooting — see the Educator Growth & Evaluation user manual, which is also linked from inside the tool.
Risk indexes, risk ratios, and composition from aggregate discipline or identification counts — the arithmetic your state uses, on counts you type in, computed entirely on this device.
First session. In Analyze, name the outcome (say, out-of-school suspensions), then enter at least two rows of counts: a group's enrollment and how many students in it had the outcome, unduplicated. Results appear as you type. Save the analysis and it joins Saved and, across years, Trends.
Keep in mind. Small groups make unstable ratios, and the tool says so rather than hiding it. When a comparison group inside the building is too small, use the alternate-comparison entry (the 34 CFR 300.647 statewide alternate risk ratio) — the tool shows the alternate ratio beneath any flagged standard one. A flagged ratio is a question to investigate, not a finding of discrimination.
Screening scores in, team-review flags out. You paste a screening window (student codes and scores), enter your benchmark and intensive cut scores, and get suggested risk bands, intervention grouping, and window-over-window movement.
First session. In Screen, describe the measure and window, set the two cut scores, and paste scores one per line as code, score. Save the window and the tier board builds. The Groups tab drafts intervention groupings you can rearrange; Progress compares windows once you have two.
Keep in mind. The tool ships no benchmark tables — publisher cut scores are proprietary and measure- and season-specific, so you must enter your own. "At the cut" deliberately counts as the less intensive band in both directions. Bands are flags for team review, never automatic placement. Progress means movement between bands, which is not the same thing as a score going up — and for measures where lower is better, the tool inverts correctly, but check the "higher is better" toggle when you set up the window.
Evaluation clocks, IEP annuals, and triennials on one urgency dashboard (overdue / due in 14 / due in 30), with per-provider caseload views. Students are entered as codes only.
First session. Add timelines on the Timelines tab; the Dashboard sorts them into urgency bands automatically.
Keep in mind — this one matters legally. Due dates are editable prefills, not legal determinations. The evaluation clock prefills the IDEA fallback of 60 calendar days from consent (34 CFR 300.301(c)); many states set their own timeframes, and several count school days, which no calendar arithmetic can compute. Confirm every date against your state rule and district calendar — the screen and the CSV export both repeat this because it is that easy to forget. Once you hand-edit a date, the prefill stops touching it.
A reference and reasoning tool: compare clinical diagnosis, IDEA eligibility, and Section 504; walk the evaluation path; review federal definitions, safeguards, and timelines; explore open-question cases; and build a privacy-safe meeting-preparation guide.
Keep in mind. It never decides eligibility, services, or placement — it prepares the humans who will. Definitions are federal; your state adds its own layer, so treat it as the map, not the territory.
Notes or a transcript in, your district's meeting format out. Six built-in formats (SST, IEP team, 504, parent/family conference, staff/PLC, evaluation planning) plus a custom-template builder — model your district's own form once and reuse it.
First session. Pick a format on New record, paste your notes, and draft. Review the result, certify it, and save; action items land in the Action items tracker with checkboxes.
Keep in mind — the two integrity mechanisms. Before any AI call, names are masked locally into "Person A / Person B" and restored locally afterward, so the model never sees who was in the room. And every drafted decision or action item carries a verbatim quote from your notes, mechanically verified; a claim the tool cannot find in your notes is flagged "not found in your notes" and kept visible rather than silently trusted. You certify before saving, and the export footer says a human did. There is also a fully manual no-AI mode.
One announcement, every family language: write the English master, pick your building's languages (sixteen presets including right-to-left scripts), translate with AI, review each translation, and export an accessible packet with every section properly language-tagged and direction-aware.
First session. Compose a short real announcement — a picture day notice is a good first run — select two or three of your building's languages, translate, and read the translations before saving. The language selection is remembered for next time.
Keep in mind. Translations are machine-assisted and you certify the final text — review with bilingual staff where you can. Every export carries a machine-translation disclosure with a contact-the-office line; do not remove it. The language tagging is not cosmetic: it is what makes the packet work in screen readers and in right-to-left scripts, which is the point of the tool.
The hub tools are ordinary AlloFlow modules: if a card opens to a loading screen that never resolves, the recovery sequence in Troubleshooting applies. AI drafting and translation require a configured backend — on a keyless install those buttons explain themselves and everything non-AI keeps working, per Privacy and responsible AI. For the rollout conversation with staff, pair this chapter with School rollout and coaching.
The STEAM Lab is the largest thing in AlloFlow: 142 interactive tools across thirteen subject areas, from a water-cycle simulation to a titration burette to a disproportionality analyzer. This chapter does not describe all 142 — that list lives in the catalog, which is always current, while a printed list would be wrong within a month. What this chapter teaches is the part that does not change: how to find the right tool, how to put it in front of students, and what to expect from any tool you open.
| Area | Roughly |
|---|---|
| Earth and space science | 27 tools |
| Computing, AI, and digital literacy | 19 |
| Ecology, environment, and animals | 14 |
| Math fundamentals | 15 |
| Life science and genetics | 12 |
| Life skills, careers, and economics | 11 |
| Engineering and design · geometry and measurement · learning and behavioral science | 7 each |
| Human body and safety · data and statistics | 5 each |
| Strategy games · sports and movement science | 4 each |
Where the live list is. All 142 live inside the STEAM Lab itself: open it from Learning Tools on the Launch Pad, or from the STEAM Lab entry in the tool list, and search or browse by area there. That is the only always-current list, which is why this chapter does not reproduce it.
Do not confuse this with Find a tool on the AlloFlow website. That page is a finder for the lesson-building tools, the ones that turn source material into glossaries, adaptations and quizzes, and it usefully filters by whether a tool needs a source, needs AI, or runs in Gemini Canvas. It does not list the STEAM Lab's simulations, so searching it for "titration" or "solar system" finds nothing.
Search, if you know roughly what you want. Open the STEAM Lab and type in its search box. It matches tool names, descriptions, topics, and hand-written aliases — so "photosynthesis" finds the tree lab even though that word is not in its title.
Browse by area, if you are planning a unit. The lab's own subject groupings are the fastest way to see everything available for a topic you are about to teach.
Ask by voice or the command palette, if your hands are busy: press Ctrl+K and type the tool name, or say it aloud with voice control on.
This is the part worth learning once, because it works for every tool.
Every tool has its own web address. A tool called Water Cycle is at /water-cycle, the coaster lab at /coaster-lab, and so on — one address per tool, all 142 of them. Paste that link into Google Classroom, an email, or a QR code and a student lands directly in that one tool, with no menus to navigate and nothing to install. They do not need an account.
A student who arrives this way sees a small note at the bottom of the screen explaining that this is one tool from AlloFlow and offering a way into the full app. It can be dismissed and does not follow them into the activity.
Or share the whole lab by sending the app link and telling students which tool to open. Use the direct link when you want them in one place; use the app link when the activity is "explore three of these."
They run on the device, in the browser. No installation, no account, and nothing uploaded. A Chromebook, an iPad, and a ten-year-old laptop all work; the 3D tools want a reasonably modern browser.
AI is optional and clearly marked. Most tools are simulations and instruments — they compute, they draw, they respond, and none of that needs AI. Some offer AI extras such as hints or coaching. If your school has not connected an AI backend, those extras are simply not shown, and a small "AI extras: off" note appears in the lab header explaining why and how to turn them on. The tool itself works fully either way. This matters for planning: a lesson built on a simulation is not at risk if the AI is unavailable that morning.
They are instruments, not answer keys. The tools show what happens; the interpretation is the lesson. A simulation that produces a surprising result is usually the most valuable moment in the class, not a bug to be fixed.
Fullscreen and accessibility. Most visual tools have a fullscreen control for projecting to a class. Tools are keyboard-operable, and the 3D ones expose camera controls to the keyboard rather than requiring a mouse drag.
As a demonstration. Project one tool, drive it yourself, and narrate. Fastest to plan, and the fullscreen control exists for exactly this.
As a station. Send the direct link for one tool to a group and give them a question to answer with it. Because each tool has its own address, four groups can be in four different tools without anyone getting lost.
As the evidence step in a lesson. Build the lesson in AlloFlow as usual, then send students to a tool to gather the observation the lesson asks them to explain. The tools do not record student work, so decide in advance how you want them to capture what they found — a worksheet, a photo, a shared document.
Almost every problem is one of three: the tool did not finish loading, the browser is too old for a 3D tool, or an AI extra is hidden because no backend is connected. The first is a reload, the second needs a different device, and the third is explained by the note in the lab header. The full recovery sequence is in Troubleshooting.
Tool links are the easiest way to hand teachers something concrete: pick three tools for a grade band and send the three addresses. There is nothing to license, provision, or roster, so a department can pilot on Monday. The privacy picture is the same as the rest of AlloFlow and is set out in Privacy and responsible AI.