Diffit + Claude: Adapting Instructional Materials for Every Student

Here is the quiet arithmetic of every classroom: one text, thirty reading levels. The article you found is perfect for the middle of the room and unreachable for the bottom third and unchallenging for the top. You know this the moment you hand it out — the kids who freeze, the kids who finish in four minutes and start talking. The standard response is to do nothing about it, because the alternative is to rewrite the same content three or four times at three or four levels of access, and there is no version of a Tuesday night with room for that. So the material meets the middle, and the two ends of the room quietly go unserved.

The Diffit connector in Claude for Teachers is built for exactly that arithmetic. Its capability names two verbs: create and adapt instructional materials for every student. Both matter, but "adapt" is the one that changes your week. Creating a new resource is a familiar job. Adapting existing material — reworking a text you already trust so it reaches every reader in the room without losing its substance — is the job you skip, every time, because by hand it's the most expensive thing you do. Making that job cheap is the whole point.

"Create" is table stakes. "Adapt" is the hard part

Plenty of tools create instructional materials. Fewer take material that's already good and make it reach further, and that second move is where the real teaching leverage is. Think about why. When you create from scratch, you get one artifact pitched at one level. When you adapt, you take a resource whose content you've already vetted — the right topic, the right rigor, the right standard — and you extend its reach across the range of students who need it. The intellectual work of choosing good material is preserved; only the access changes.

That distinction is the difference between differentiation and dilution. The failure mode teachers rightly fear is "adapting" that quietly strips the concept down until the struggling readers are working on something easier rather than the same thing made reachable. Real adaptation holds the core content steady and changes the access to it — the reading level, the scaffolding, the entry point — so that every student is wrestling with the same idea at the level of challenge that's productive for them. Diffit adapting for every student is a claim about reach, not about lowering the bar.

Every classroom is a range of readers wearing one label

We organize schools by grade, which encourages a fiction: that a class of "seventh graders" reads at a seventh-grade level. It never does. A single roster spans years of reading ability, plus multilingual learners, plus students with specific needs, plus the kids who are simply having a hard month. The grade label is an administrative convenience, not a description of the readers.

Instruction that ignores that range doesn't serve everyone equally — it serves the middle and leaves the edges to fend for themselves. The struggling reader can't access the text, so they can't access the content, so the gap between them and the concept isn't a learning gap at all; it's a decoding gap dressed up as one. And the advanced reader, handed material well below their level, learns to coast. Adapting material for every student is how you attack both problems at once: same concept, right level of access, no one stuck at the door and no one bored inside it.

Differentiation isn't about giving different students different content. It's about giving them the same content through a door each of them can actually walk through. Diffit's job is to build the extra doors — fast enough that you'll actually use them.

Adaptation is the highest-value task, which is why it's the first one skipped

Ask teachers to rank instructional practices by impact and differentiation sits near the top; ask them what they actually have time to do and it sits near the bottom. That gap — high value, low feasibility — is precisely the gap Claude for Teachers was designed to close. The research on what reliably helps students (differentiation, mastery-based learning, small-group instruction) has been settled for a long time. The obstacle was never knowledge. It was that a teacher's week doesn't contain the hours these practices demand by hand.

Diffit attacks the feasibility side of that equation. When adapting a text for every reader in the room costs seconds instead of an evening, the practice that always got skipped becomes the practice you can actually run — Monday, and again Tuesday. That's the mechanism by which "best practice" stops being a phrase in a professional-development binder and becomes something that happens in the room:

• The value was never in doubt. Meeting students where they are is proven pedagogy; nobody's arguing the point.
• The cost was the whole problem. By hand, adapting one resource for four readiness levels is an evening you don't have, so it doesn't happen.
• Cheap adaptation changes the default. When the reworked version takes seconds, the struggling reader gets the accessible text as a matter of course, not as a heroic exception.
• The core content holds. Because you're adapting trusted material rather than generating new content per student, everyone stays on the same standard — the access changes, the substance doesn't.

The connector doesn't make differentiation a good idea. It already was. The connector makes it affordable, and affordability is the only thing that was ever standing between the idea and the classroom.

Diffit and the built-in differentiation workflow

Claude for Teachers already ships a native way to differentiate. The open-source k12-lesson-differentiation skill adapts an existing lesson into tiered versions — below, at, and above proficiency level — keeping the core content consistent across tiers and producing a differentiation plan plus per-proficiency student materials. If you want to see that mechanic in depth, One Lesson, Three Readiness Levels walks through it.

So where does the Diffit connector fit next to a built-in skill that already tiers lessons? Think of them as two reaches at the same goal. The differentiation skill produces a structured, tiered plan grounded in the Learning Commons Knowledge Graph standards; the Diffit connector is a dedicated surface for creating and adapting instructional materials for every student. Reach for the skill when you want the full differentiation plan with per-proficiency materials as a unit; reach for the connector when you have a specific resource to adapt for the range in front of you. Both hold the core content steady; both exist so no reader is left at the door.

They also compose with the wider ecosystem. A connector like MagicSchool makes instructional content classroom-ready; Diffit adapts that content for every student; the differentiation skill structures the whole thing into tiers. Create, adapt, tier — a clean division of labor, each piece owning one move. If you want a repeatable scaffold around them, a marketplace bundle of education agent skills or a structured lesson plan studio keeps the upstream planning organized enough that the adaptation has good material to work from.

And the barrier to trying it is gone. Claude for Teachers is entirely free for verified US K-12 educators, with sign-up open through June 30, 2027 for a full year of access. Which means the two students at the edges of your room — the one who can't get in the door and the one who's bored once inside — finally get material built for them, not the fiction that they're the same reader as everyone else.

Part of the Claude for Teachers series. Related: One Lesson, Three Readiness Levels: The Differentiation Problem, Solved · MagicSchool + Claude: Making Instructional Content Classroom-Ready. Browse AI tutoring skills or more builder insights.