Kindergarten suite · everything proposed so far
The Kindergarten Tool Index
Every tool across the three playbooks in one list, with what is actually built, what has a
design behind it, and what is still only a word on a page.
Built running and usable
Specified a playbook says how
Named only no design yet
01What exists
Two tools out of roughly fifty ideas. Both came out of the same shell, which is the argument
for the shell.
Sounds I Can Make
RF.K.4.a · RF.K.4.b
Built
Letter-sound probe with Letterland characters, a hashed student number so no name is stored,
and a scratch-off reward. Covers consonant sounds and long and short vowels.
letterland-sounds.vercel.app
Build the Number
NC.K.OA.1 · NC.K.CC.3
Built
One number shown six ways at once. Three tiers — find it, build it, your way —
with the next tier earned at five stars. Records which representations a child misses and,
crucially, which square they tried to drop the card into, then offers a three-number drill
across just the two they keep confusing.
build-the-number.vercel.app · teacher passcode at /pin
02Mathematics — eighteen tools
Six categories, from Kindergarten Math at the Board. Every one of these is a
Promethean tool: lower half of the glass, palm-sized targets, several children at once.
A · Count and quantify CC.1 · CC.2 · CC.4 · CC.5
- Hundred chart that lights as the class chantsSpecified
Counting to 100 by ones and tens, with the pattern visible rather than recited blind.
- Tap-to-count setsSpecified
A counted object changes and stays changed, so double-counting is impossible by design instead of corrected after.
- Number-line hops from a given startSpecified
Counting on from seven, rather than starting again at one.
B · See it without counting CC.4 · OA.5
- Flash-and-hide dot patternsSpecified
The only way to force subitising. The limit is on the stimulus, not on the child's answer, which is why it is the one timed thing allowed at a board.
- Combine two groupsSpecified
Seeing six as four and two — conceptual subitising, which is the harder half.
- Pattern strips with a gap to fillSpecified
Worth teaching. Not an NC kindergarten standard, so not worth reporting on.
C · Represent and translate CC.3 · OA.1
- Build-the-number matBuilt
The connection half of OA.1, which worksheets almost never ask for.
- Sort mixed representations into numbered binsSpecified
Fast, forgiving, and works with four children at once.
- Match the pairsSpecified
Memory, played with a tally against a ten frame rather than two identical cards.
D · Compose and decompose OA.3 · OA.4 · NBT.1
- Pull-apart number bondSpecified
Decomposition as a physical act. The total visibly stays the same while the parts move.
- Double ten frame for teensSpecified
Fourteen as a full ten and four more, which is the whole of NBT.1.
- Partners-of-ten machineSpecified
Give it six, it wants four. One interaction, ten facts.
E · Compare CC.5 · CC.6 · CC.7
- Matching bridge between two setsSpecified
Draw lines pairing two groups until one runs out — greater and less exactly as CC.6 words it.
- One-more conveyorSpecified
An object arrives, the child names the next number. CC.5 in one gesture.
- Numeral comparison, no objectsSpecified
CC.7 is deliberately harder: two written numerals and nothing to count.
F · Form and write CC.3 · and RF.K.2 over in ELA
- Numeral tracing with start point and directionSpecified
Formation, not just shape. Where the pencil starts is most of the lesson.
- Line and curve pathsSpecified
Fine-motor readiness. Not a maths standard at all — its nearest home is handwriting.
- Write then check against a setSpecified
Keeps the quantity attached so it never becomes pure handwriting.
The eleven interaction patterns underneath them
Build these once and the maths suite is mostly configuration. Three of the eleven are done.
| Pattern | What the child does | State |
| Build-the-number | Fills empty slots with representations of one number | Built |
| Sort into bins | Throws mixed representations into numbered bins | Built as the drag half of the mat |
| Tap-to-count | Touches each object once; it changes and stays changed | Built as the tier-two counter |
| Drag-to-frame | Pulls counters into a ten frame or a set ring | Specified |
| Pull-apart | Drags a whole into two parts, both numbers moving | Specified |
| Flash-and-hide | Sees a dot pattern for a second, then it is gone | Specified |
| Trace | Follows a numeral path with a finger from the right start | Specified |
| Number-line hop | Jumps a marker forward from any starting number | Specified |
| Matching bridge | Draws lines pairing two sets until one runs out | Specified |
| Fill the gap | Completes a repeating pattern from a tray | Specified |
| Parallel lanes | Two to four children, same task, side by side, at once | Named only |
Parallel lanes is the one that changes the economics. It is not content, it
is a wrapper, and without it a Promethean panel runs one child at a time while nineteen
watch. It is also the only pattern in the list with no design written for it.
03Reading foundations — eleven pieces
From Mapping the Alphabet Code. These are not separate apps so much as the rest of the
probe that already exists, plus the engine that makes it adaptive.
Content still missing from the probe
- Phoneme blendingSpecified
The hidden gate. A child can know all twenty-six sounds and still not read a word. Oral, no print in front of them — and it changes how every letter-sound result already collected should be read.
- Letter names, all 52 formsSpecified
Same mechanic as sounds, doubles the coverage, and lets names and sounds be told apart as separate problems.
- Rhyme, syllables, onset and rimeSpecified
Where a child goes when print is premature. All oral.
- High-frequency wordsSpecified
A parallel track, not downstream of decoding.
- Handwriting, upper and lower caseSpecified
RF.K.2 is North Carolina's own cluster, with no Common Core equivalent. It is why every foundational code after it shifts by one.
The engine
- Item state machineSpecified
Unknown, Learning, Review, Secure, with spacing doing the work. Three right in one sitting is short-term memory; three across three weeks is learning.
- Twelve-item session budgetSpecified
Seven at the edge, three maintaining, two already secure. The last two are not sentiment — a probe that is all failure teaches a child that reading is where they are bad at things.
- Confusion captureSpecified
Record what they said, not just whether it was right. One extra tap, the largest single gain in diagnostic value. The same idea now running in Build the Number.
- Dependency gateSpecified
Refuse to schedule a digraph whose parts are not secure. Teaching sh before s and h wastes the lesson.
- Entry, ceiling and floor rulesSpecified
Start a tier below the last secure skill. Five right, move up; three wrong, drop down. Never keep going to get a complete picture.
- Server, accounts, class reportingSpecified
Deliberately last. Everything above teaches you what needs storing.
04The rest of ELA — six capabilities
From The Kindergarten Playbook. Organised by what the software does rather than which
standard it touches, because a kindergartener cannot read instructions or type an answer, and
that single fact disqualifies most of what gets sold into primary classrooms.
Across all 39 live standards
- Capture — hold to recordSpecified
Fifteen seconds of audio tagged to a child and a standard. No scoring, no transcription. The teacher already made the judgement; the software just stops it evaporating. Serves about 24 standards and is a week of work. Hardest part is being faster than a sticky note.
- Capture — photograph a drawingSpecified
One photo becomes evidence for a writing standard and a speaking standard at the same time.
- Draw, dictate, transcribeSpecified
Canvas, microphone, live transcription. A real writing instrument for a child with no spelling — and watching your own words become letters is itself a print-concepts lesson. W.K.4 is the one standard that cannot be met without software.
- Decodable text matchingSpecified
Given the fourteen sounds this child has been taught, which books in the room can they read without guessing? Tedious and exact, which is the definition of work to give a machine.
- Grouping by shared gapSpecified
Small groups that form around a lesson rather than a label.
- Book-to-standard lookupSpecified
Which text on the shelf gives a clean shot at this week's standard.
- Facilitation oddmentsSpecified
Turn-taking tokens, a visible talk timer, partner pairing that does not repeat the same two children, sentence stems shown as pictures. Software as the object on the table, not something a child operates. Small, cheap, used daily.
- Coverage gridSpecified
Thirty-nine standards by twenty children, coloured by how old the evidence is. The holes are the product. Nearly free once capture exists, and it is the thing that makes a teacher open the app on a Sunday.
- The quiet-child alarmSpecified
Who has the least evidence this fortnight. Usually the child who most needs looking at.
- Report assemblySpecified
Pull one child's captured evidence into a draft a human then edits. 780 judgements a reporting period; the exhausting part is not making them.
- Conferring notesSpecified
Attach to a standard as they are typed, so filing is free.
Read the capability list rather than the standard list. Capture and teacher
support are the only two that serve every strand. Discrete practice — the thing most
people picture when they hear educational software, and the only thing built so far —
serves exactly one.
05What has no playbook
These came up in conversation and never got written down anywhere. Listing them honestly
matters more than padding the sections above, because a tool with no design behind it is not
nearly built.
ELA sub-skills named but not designed
- First sound · last soundNamed only
Phoneme isolation. Sits beside blending in the foundational strand and would reuse the same oral shell, but no item design exists.
- Author and illustrator, and what each doesNamed only
A literature standard. No interaction proposed, and the playbook argues software should not try to score this family at all.
- PunctuationNamed only
A language-conventions standard. Plausibly a discrete-practice item type, which would make it one of the few in that strand that is.
- Sentence writingNamed only
Probably a mode of the draw-and-dictate tool rather than its own thing.
- Labelling a drawingNamed only
Same tool, different prompt. Cheap if draw-and-dictate exists, nothing without it.
- Sequencing a storyNamed only
Drag four pictures into order is the obvious build, and also the one most likely to reward a child who remembers the pictures rather than the story.
Whole subjects named as categories
- Social StudiesNamed only
Named as one of the four subjects. No standards pulled, no categories, no tools.
- ScienceNamed only
Same. Both would need their own DPI documents read before anything is designed, and neither has been.
So: two built, about forty with a design
written for them, eight that are a phrase and nothing more, and
two whole subjects not started.
06The rules that apply to all of it
Never score a retell, or judge a conversation. Keyword matching rewards a
child who says the right nouns in the wrong order. There is no signal in a microphone for
whether someone built on what a classmate said.
Never produce a reading level from a probe. Ten letter sounds is not a
Lexile, and the moment a number looks official it ends up in a file that follows the child.
Never put scores on the glass. Twenty children can read who got it wrong.
Judgement belongs on the teacher's device.
Never time a whole class. The slowest child at the board is being watched
by nineteen others.
Never drag to a target smaller than a palm. A five-year-old at arm's length
on glass has about an inch of accuracy. Snap generously, or do not drag.
Never replace the adult listening. Most of these standards contain the
phrase “with prompting and support”. The prompting is the standard.
Board and tablet are different jobs. At the board you are teaching: whole
class, several hands, and what you can capture is who came up and what happened — a
note, not a score. On a tablet beside one child you are assessing, and the record is real.
A board tool that claims to have assessed twenty children has assessed nobody.
The test for any of it: does this save a teacher time, or spend a child's?
Kindergarten has roughly nine hundred instructional hours in a year.
07If the next thing matters most
Three build orders exist, one per playbook, and they disagree because each was written for its
own subject. Merged, and judged on good done per week of work:
- Capture. One screen, a class list, the standards, hold to record. A week
of work, two dozen standards on day one, and it is the only thing on this page that helps
with every strand.
- The coverage grid on top of it. Nearly free once capture exists.
- Blending in the letter-sound probe. It is the gate, it is cheap to test
orally, and it changes the meaning of every result already collected.
- Pull-apart number bond. One interaction covering three maths standards.
- Draw and dictate. Satisfying for children, and the only route to the one
standard that requires software.
- Parallel lanes around whichever board tool is working best — at
which point the panel starts earning its price.
Compiled from Mapping the Alphabet Code, The Kindergarten Playbook and
Kindergarten Math at the Board. Standard codes in sections 02 to 04 are taken from
those documents, which cite the NC Standard Course of Study for ELA and Mathematics, adopted
2017. Section 05 deliberately carries no codes: nothing in it has been checked against a DPI
document.