Education Agent Skills

A free, open library that teaches AI systems how good teachers actually work — how they give feedback, ask questions, scaffold, and assess — so AI tools in classrooms act with pedagogical judgement instead of just answering.

What should AI agents know how to do before they touch education?

An open library of educator skills for human and machine use: feedback, questioning, scaffolding, agency, assessment, and learning design.

Tool access is cheap. Pedagogical judgement is the scarce layer.

Open Agent Skills →All work

Education Agent Skills

20 domains. 165 skills.

AI & Learning Science14

AI Literacy7

Curriculum Alignment4

Curriculum & Assessment13

EAL & Language Development5

Environmental & Experiential Learning6

Explicit Instruction5

Global & Cross-Cultural Pedagogies9

Historical Thinking10

Inclusive Design3

Literacy & Critical Thinking7

Memory & Learning Science8

Montessori & Alternative Approaches4

Original Frameworks17

Professional Learning10

Questioning & Discussion5

Self-Regulated Learning5

Student-Facing Learning Skills13

Systems Thinking8

Wellbeing, Motivation & Agency12

What it is

An open library of educator skills for human and machine use: 20 domains, 165 skills covering feedback, questioning, scaffolding, agency, assessment, and learning design.

Why it exists

Tool access is cheap; pedagogical judgement is the scarce layer. The library puts evidence-based educator judgement into a form both teachers and AI agents can use.

What it contains

Twenty domains — from AI and learning science to historical thinking, self-regulated learning, and wellbeing — each holding named skills, 165 in total. A hosted MCP server makes the library available to AI agents.

Current status

Open source, public on GitHub.