Skip to content
RosterRoster

The library

Every move, in one place.

117 free lessons. The whole week, planning through parent night, plus building your own tools.

Tap one, or describe what you're stuck on below.

New here? Start with Compliance & safe AI use (free) and go in order. Jump anywhere once it clicks. The Builder track is its own path. Start it whenever you want AI building FOR you.

✎ Ask Roster for a lesson on anything

Describe a real thing you need to get done. Roster builds the drill from your exact situation.

The classroom work

Planning to feedback to families. 12 areas · 32 lessons on the classroom work, all free.

1Compliance & safe AI use (free)7 lessons›

No student goes in by name, grounding, and the decision that stays yours. The skeleton, practiced to reflex.

Show it one unit plan that cleared the chair›What student information can go in a prompt (almost nothing)›Build your course's planning master prompt›Delegate a full pre-send review›Make AI defer to your records, not its guess›Force the answer into a decision matrix›Convene an AI board of advisors›
2Workflows & automation5 lessons›

Design a repeatable task into an AI workflow. With a human approval gate exactly where it belongs.

Map a repeatable task into a workflow›Design the approval gate before anything goes home›Spec the follow-up cadence for missing field-trip forms›Automate the unit-launch checklist›Set the control dial: where a human stays in the family-communication workflow›
3Claude & Cowork mastery3 lessons›

One platform, mapped to the job: a Project for your standing context, a Cowork skill for a repeatable prompt, MCP connectors for your live inbox and calendar.

Give Claude your course once›Turn your best prompt into a skill›Let Claude read your inbox and calendar›
4Student data and the line you cannot cross6 lessons›

What an education record is and why the model never needs one, taking every student out of a paste in thirty seconds, why the grade and the placement stay yours, your district's policy read as a list, the approved-tool list against the vendor's promise, and the under-13 line.

What an education record is, and why the model never needs one›Take every student out of a paste in thirty seconds›The decision stays yours: grades, placements and consequences›Read your district's AI policy before you read any lesson here›The approved-tool list, and what a vendor's no-training promise does not settle›Under thirteen: the line that moves when the students do›
5Planning6 lessons›

A unit from a standard with the standard open, one lesson at three levels, the sub plan at six in the morning, the rubric before the assignment, backward design in one pass, and adapting a plan you did not write.

A unit plan from a standard, with the standard open beside it›One lesson, three versions, same objective›The sub plan at six in the morning›Write the rubric before the assignment›Backward design in one pass: assessment, evidence, then the days›Adapt a plan you did not write without losing what it was for›
6The teacher's week5 lessons›

Audit the week before you hand the first task over, catch the standard quoted with one clause changed, cost the scripted run that needs the roster, wrap the day into what is still your call, and know where a paste goes when the student was in it.

Audit your week before you hand the first task over›Audit an answer before it reaches a student or a family›Cost a scripted run before you build it›Wrap the day: what went home, what is still yours›The rules, and where a paste goes when the student was in it›
7Feedback and assessment0 lessons›

Rubric feedback that quotes the work, the comment bank that stays in your voice, the exit-ticket sort, the test that matches the unit, and the grade that is never the tool's. Written next.

Rubric feedback that quotes the work, with the grade left as your callSoon
The comment bank in your own voice, built from your last thirty commentsSoon
The exit-ticket sort: three piles, no names, tomorrow's first five minutesSoon
A test that matches the unit, item by item against the standardSoon
Feedback on thirty essays that each gets its own sentenceSoon
The grade is never the tool's: the audit before the gradebookSoon
8Families and communication0 lessons›

The email home that a parent can act on, the newsletter in twenty minutes, the conference prep sheet, the hard conversation drafted and then rewritten by you, and translation you can check. Written next.

The email home a family can act on, student lettered until you add the nameSoon
The class newsletter in twenty minutes, from the week's planSoon
The conference prep sheet: three strengths, one ask, one questionSoon
The hard conversation, drafted, then rewritten in your own wordsSoon
Translation you can check: back-translate before it goes homeSoon
What families are entitled to know about the toolSoon
9IEPs, 504s and support0 lessons›

The present-levels draft from the data you have, accommodations in the lesson plan, the progress note the team can read, the meeting agenda, and what never goes in a tool at all. Written next.

Present levels from the data you already have, every judgment marked as the team'sSoon
Accommodations written into the lesson plan, not stapled to itSoon
The progress note the team can read in a minuteSoon
The meeting agenda that keeps the hour on the studentSoon
What never goes in a tool at all: the evaluation, the eligibility, the placementSoon
The case manager's week, auditedSoon
10Classroom and behavior0 lessons›

The routine written out for a substitute, the behavior plan as a recipe, the incident write-up with the student lettered, restorative questions, and the seating chart logic you own. Written next.

The routine, written out so a stranger can run the roomSoon
The behavior plan as a recipe, with the consequence left as your callSoon
The incident write-up, student lettered, facts onlySoon
Restorative questions for a conversation you leadSoon
The seating chart logic you own; the tool never sees the chartSoon
The call home after a hard day, drafted and then made yoursSoon
11Your own time0 lessons›

The weekly plan in one paste, the PD reflection, the evaluation portfolio, the resume that names what you built, and the boundary on the evening. Written next.

The weekly plan in one paste, from the unit and the calendarSoon
The PD reflection that says what changed in the roomSoon
The evaluation portfolio, assembled from what you already wroteSoon
The resume that names what you built, not what you were assignedSoon
The boundary on the evening: what the tool takes and what it does notSoon
The summer plan, one paragraphSoon
12Teaching AI, and teaching with it0 lessons›

The AI-use policy for your classroom, the assignment a model cannot do for a student, teaching the audit, the under-13 unit, and what to tell families. Written next.

The AI-use policy for your classroom, one page, inside the district'sSoon
The assignment a model cannot do for a studentSoon
Teaching the audit: students list the claims and open the sourcesSoon
The under-13 unit, with the teacher at the keyboardSoon
What to tell families about AI in the roomSoon
The AI-literacy standard, where your state has one, read as a listSoon

Builder track

Stop using AI. Start building with it. Claude Code, skills, agents, and systems that run without you. 13 areas · 85 lessons, capability-first, all free.

B1Setup & first contactInstall Claude Code, meet the terminal, and have your first real conversation about a unit, with no student in it.›
Four words before you startBuild →Pick a plan & installBuild →Terminal tour: read your dashboardBuild →IDE setup: give it a windowBuild →Your first real task, iteratedBuild →Your files as contextBuild →Talk to it: voice inputBuild →
B2The project brainCLAUDE.md sets the trajectory: your grade, your standards, your district's approved tools, and the never-lines a department chair would hold you to.›
CLAUDE.md: set the trajectoryBuild →/init on a real folderBuild →The 200–500 line ruleBuild →Split into rules/Build →Auto-memoryBuild →Secrets & .env hygieneBuild →
B3Steering & safetyA minute of planning saves ten of building, and the gate before anything reaches a family or a file is you.›
Permission modesBuild →Grill, then planBuild →/permissions rulesBuild →Bypass mode, safelyBuild →Cost capsBuild →The human-in-the-loop matrixBuild →
B4Context engineeringKnow what fills the window before you paste the whole unit; a standard, a rubric and a week fit, a gradebook never goes in.›
/context anatomyBuild →Handoff, clear, compactBuild →Token densityBuild →Specific beats vagueBuild →Pick model & effortBuild →The overhead auditBuild →
B5The verify loopTask, do, verify. A worksheet with one wrong answer in the key goes to thirty children; the verify step is the answer key, worked.›
Task → do → verifyBuild →Screenshot loopsBuild →Tests as verificationBuild →Fresh-data testingBuild →Cold-start runsBuild →Self-annealingBuild →
B6Skills: package your processYour repeatable routines become Skills: the sub plan, the rubric, the redactor that runs first every time.›
SOP → SKILL.mdBuild →Frontmatter anatomyBuild →Scripts as musiciansBuild →Iterate to 99%Build →Self-healing skillsBuild →Your skill libraryBuild →
B7Connect your tools (CRM, calendar, your data)Connect the calendar and the standards document, never the student information system; what a connector can see is what the tool can see.›
What MCP isBuild →Install & vet a serverBuild →Audit the token costBuild →Paste the API docs insteadBuild →MCP → skill conversionBuild →Build a tiny serverBuild →
B8The DO frameworkCode executes, the LLM judges. The worksheet set from a standard is code; the grade is neither, it is yours.›
Docs → chatbots → agentsBuild →The PTMRO loopBuild →Directive anatomyBuild →Code executes, the LLM judgesBuild →Compound-error mathBuild →MetadirectivesBuild →
B9Subagents & parallel workFan out when it pays: three versions of one lesson in parallel, and a reviewer that checks each against the one exit ticket.›
When to fan outBuild →Reviewer & QA agentsBuild →Summaries, not raw textBuild →Git worktreesBuild →3 approaches in parallelBuild →The 3–4 agent capBuild →
B10Design: flyers & social graphicsReference-driven design for the parent-night handout and the classroom poster, with no student's name or face in any of it.›
The screenshot-compare loopBuild →Reference-driven designBuild →Component librariesBuild →Decks & one-pagersBuild →Brand consistencyBuild →The design QA criticBuild →
B11Office & data agentsSpreadsheets, documents and reports: the unit calendar, the materials list, the weekly plan, never the gradebook export.›
Excel that explains itselfBuild →docx & pptx skillsBuild →Sheet-connected pipelinesBuild →The weekly market report factoryBuild →Ask-your-workbook agentBuild →Deliverable factoriesBuild →Build your market second brainBuild →
B12Ship a real toolShip a real tool: the sub-plan folder that builds itself, the weekly worksheet set on a schedule, with a stop rule on any file that carries a name.›
Deploy a static siteBuild →Wrap a skill in a URLBuild →WebhooksBuild →Cron in plain EnglishBuild →Secrets & observabilityBuild →Design pass: ten checks on a phoneBuild →Capstone: ship your classroom toolBuild →One real build, start to finishBuild →
B16The work folderFrom chat to a folder of your own files that one agent reads and writes. Keep a list of your corrections, turn it into a skill, shape the folder around the work, and hand it on.›
Which layer are you onBuild →Test your setup with one tiny jobBuild →Keep a list of your correctionsBuild →Turn corrections into a skillBuild →Start with the outcomeBuild →Shape the folder around the workBuild →Give each stage a fileBuild →One model, different jobsBuild →Keep it current and hand it onBuild →

Vendor notes, dated

Prices and plan names move. Each note says when it was looked up, whether a person has confirmed it, and the day it comes due. Past that day, treat it as a question.

  • Claude. Claude Pro is $20 a month billed monthly, or $17 a month billed annually. Researched Oct 6, 2026, not yet confirmed by a person, recheck by Dec 15, 2026.
  • ChatGPT. ChatGPT Plus is $20 a month. Researched Oct 6, 2026, not yet confirmed by a person, recheck by Dec 15, 2026.
  • Google Gemini. Google Gemini's paid consumer plan is Google AI Pro at $19.99 a month; Google AI Ultra starts at $99.99 a month. Researched Oct 6, 2026, not yet confirmed by a person, recheck by Dec 15, 2026.
  • Anthropic. Anthropic's consumer plans (Free, Pro, Max) may use chats to train models under a setting the user controls, with five-year retention when training is on and 30 days when it is off; Claude for Work, Education, Government and the API are excluded. [P] Researched Oct 6, 2026, not yet confirmed by a person, recheck by Dec 15, 2026.
  • Anthropic. Anthropic's usage policy, effective 2025-09-15, requires a qualified human reviewer and disclosure of AI use for housing, legal and financial decisions. [P] Researched Oct 6, 2026, not yet confirmed by a person, recheck by Dec 15, 2026.
  • OpenAI. OpenAI's consumer plans train on chats unless the Improve the model setting is turned off, and a Temporary Chat is not used for training; Enterprise, Business, Edu and API data are not used by default. [S] Researched Oct 6, 2026, not yet confirmed by a person, recheck by Dec 15, 2026.
  • Google Gemini. Google Gemini Apps keeps activity 18 months by default, keeps human-reviewed chats up to three years, and tells users not to enter confidential information. [P] Researched Oct 6, 2026, not yet confirmed by a person, recheck by Dec 15, 2026.
  • Microsoft Copilot. Microsoft Copilot on a personal account may use chats to train models unless the user opts out; Entra and Microsoft 365 work accounts are excluded. [S] Researched Oct 6, 2026, not yet confirmed by a person, recheck by Dec 15, 2026.
  • Claude. Claude Team, Enterprise and API inputs and outputs are not used for training by default; a BAA is available only on Enterprise and the first-party API after admin action. Researched Oct 6, 2026, not yet confirmed by a person, recheck by Dec 15, 2026.
  • Claude. Claude Free, Pro and Max inputs can be used for training if the user allows model improvement, with retention up to five years if allowed and 30-day deletion otherwise (pages dated 2026-03-16 and 2026-07-01). Researched Oct 6, 2026, not yet confirmed by a person, recheck by Dec 15, 2026.
  • OpenAI. OpenAI API inputs are not used for training by default; abuse-monitoring logs are kept up to 30 days; zero data retention is available by approval. Researched Oct 6, 2026, not yet confirmed by a person, recheck by Dec 15, 2026.
  • OpenAI. OpenAI Business, Enterprise and Edu no-training defaults are reported by secondary sources only; the vendor trust pages returned 403 on the check date. Read the vendor page yourself. Researched Oct 6, 2026, not yet confirmed by a person, recheck by Dec 15, 2026.
  • Microsoft. Microsoft Copilot and Copilot Chat with Enterprise Data Protection: prompts and responses are not used to train foundation models, the DPA applies, and Microsoft acts as processor (Learn page updated 2026-08-18). Researched Oct 6, 2026, not yet confirmed by a person, recheck by Dec 15, 2026.
  • Google. Gemini in Google Workspace content is not human-reviewed or used for training outside the customer's domain; default retention is 18 months. Researched Oct 6, 2026, not yet confirmed by a person, recheck by Dec 15, 2026.
  • Every vendor. Every vendor's no-training or no-retention term is a contract term, not the determination: it does not by itself make the vendor a school official with a legitimate educational interest under 34 C.F.R. 99.31(a)(1), which the district makes in its own agreement, and it says nothing about whether the plan a teacher actually uses is the plan the term covers. Researched Oct 6, 2026, not yet confirmed by a person, recheck by Dec 15, 2026.

Every area gives away its first moves free. Each track spins endless personalized lessons.