Claude for Teachers adds lesson prep and math checks built with Learning Commons

Anthropic and Learning Commons have co-developed two teaching skills grounded in learning science, as Claude for Teachers expands to schools and districts as a free enterprise offering

Anthropic and Learning Commons have added lesson preparation and math Check for Understanding skills to Claude for Teachers

Anthropic has added two new teaching skills to Claude for Teachers, giving educators structured support for preparing lessons and checking students’ mathematical understanding.

The skills were co-developed with Learning Commons and focus on two specific classroom workflows: Lesson Preparation and Check for Understanding.

Lesson Preparation is designed to help teachers work through an upcoming lesson before teaching it, including identifying key instructional moments and anticipating where students may struggle.

Check for Understanding takes a narrower assessment role. It generates a short math check, typically one to three items, intended to surface what students understand about a particular standard and give teachers information they can use when deciding what to teach next.

Learning Commons says the latter was refined through evaluation and feedback from Student Achievement Partners.

The additions arrive alongside Anthropic’s expansion of Claude for Teachers as a free enterprise offering for schools and districts, moving the product beyond individual teacher use while adding more structured instructional workflows.

Sandra Liu Huang, President of Learning Commons, said on LinkedIn that the two organizations had built the skills “grounded in learning science and designed to support high-quality instruction.”

From general AI assistant to defined teacher workflows

The approach is more structured than simply giving teachers a general-purpose chatbot and asking them to prompt it for a lesson plan.

Learning Commons describes its Agent Skills as packages containing the instructions, references and guardrails an AI assistant needs to complete a particular K-12 teacher workflow more consistently.

Its wider open repository currently includes skills for checking understanding, differentiating lessons, creating lesson plans and preparing to teach an existing lesson.

The two skills incorporated into Claude for Teachers are Lesson Preparation and Check for Understanding.

Learning Commons says its skills draw on its Knowledge Graph and are intended to produce materials aligned with pedagogy and academic standards. The Knowledge Graph connection is recommended rather than mandatory in the wider open implementation.

Learning Commons has made its Agent Skills cross-platform and model-agnostic, rather than restricting them to Claude.

The open skills can be installed in agent environments that support the format, with Learning Commons listing Claude Code, Cursor and Codex as examples. Developers can also use the skills and its MCP server through APIs when building them into other products.

That means the Anthropic collaboration is one implementation of a broader set of education-focused AI components, rather than a closed collection of workflows available only through Claude.

Math checks are designed to expose what students understand

The Check for Understanding skill is particularly specific about what the AI is supposed to produce.

Instead of generating a longer worksheet or generic question bank, it creates a short assessment focused on a particular math standard. Learning Commons positions the output as a way of identifying a student’s current understanding so that subsequent teaching can respond to it.

Lesson Preparation works differently. Rather than generating an entirely new lesson, it is designed to help a teacher internalize an existing one through conversation, including anticipating likely difficulties and preparing responses before teaching.

Lesson Plan Creation, which builds standards-aligned lesson plans and can optionally align them to a teacher’s curriculum.

The organization also publishes evaluation rubrics alongside the skills covering areas including pedagogy, rigor, formatting and model scaffolding. Those rubrics can be used to assess generated materials.

Learning Commons does not guarantee that outputs will be accurate or suitable in every context. Its repository explicitly states that the resources are provided “as-is” and that users remain responsible for evaluating whether generated materials are appropriate for their particular use case.

Open infrastructure sits behind the Claude partnership

Learning Commons positions the work as part of a wider effort to create reusable infrastructure for education AI rather than develop isolated teacher tools.

Huang said the Anthropic collaboration reflects a goal of building “open, public infrastructure for a new generation of education technology tools that are grounded in pedagogy, aligned to state academic standards, and designed for the common good.”

The organization is also inviting other partners to co-develop K-12 workflows and says it can provide early access to new skills, tools and its Knowledge Graph.