GarethManningGarethManningCommunityΒ·114 Agent Skills Included

claude-education-skills

Evidence-based teaching skills for lesson design, assessment, and AI tutoring

Generates research-grounded lesson plans, assessments, rubrics, tutoring dialogues, and curriculum frameworks for teachers and education designers. Replaces hours of manual research and planning with structured, evidence-cited outputs across 20 education domains. Helps educators, curriculum designers, and EdTech builders produce classroom-ready materials grounded in named learning science.
npx skills add GarethManning/claude-education-skills --all -g -y
Available:

Tells the AI agent how the 165 skills are organised, which files to read at session start, and the mandatory rebuild steps required after editing any skill so changes reach the live MCP server.

All Skills in This Repository (114)

Pure Emerald Level Indicators
πŸ“¦ In Repo
GarethManningGarethManning

formative-assessment-loop-designer

Designs inner- and outer-loop formative assessments for AI learning environments.

Community
Advanced
πŸ“¦ In Repo
GarethManningGarethManning

erroneous-example-designer

Design three erroneous worked examples with one realistic error per domain.

Community
Advanced
πŸ“¦ In Repo
GarethManningGarethManning

worked-example-to-problem-solving-transition-designer

Design adaptive fading transitions from worked examples to independent problem solving.

Community
Advanced
πŸ“¦ In Repo
GarethManningGarethManning

intelligent-tutoring-dialogue-designer

Design phase-based tutoring dialogue frameworks that elicit student reasoning before explanations.

Community
Advanced
πŸ“¦ In Repo
GarethManningGarethManning

digital-worked-example-sequence

Design digital worked example sequences with fading and self-explanation prompts.

Community
Intermediate
πŸ“¦ In Repo
GarethManningGarethManning

productive-failure-desirable-difficulty-designer

Redesign teaching sequences to begin with generation and structured consolidation.

Community
Advanced
πŸ“¦ In Repo
GarethManningGarethManning

metacognitive-monitoring-ai-contexts

Diagnose metacognitive risks in AI-assisted learning and design calibration-focused monitoring interventions.

Community
Advanced
πŸ“¦ In Repo
GarethManningGarethManning

cognitive-tutoring-architecture-designer

Decompose target skills into testable knowledge components with Bayesian knowledge tracing.

Community
Advanced
πŸ“¦ In Repo
GarethManningGarethManning

adaptive-hint-sequence-designer

Generate a cascading hint sequence for a problem type and sticking points.

Community
Advanced
πŸ“¦ In Repo
GarethManningGarethManning

ai-facilitated-collaborative-learning-designer

Design AI-supported collaborative tasks with phases, roles, and regulation scaffolds.

Community
Advanced
πŸ“¦ In Repo
GarethManningGarethManning

learning-analytics-interpretation-guide

Identify actionable teaching decisions from LMS dashboards, quiz patterns, and engagement metrics.

Community
Intermediate
πŸ“¦ In Repo
GarethManningGarethManning

self-explanation-prompt-designer

Design self-explanation prompts with rationale, quality indicators, and scaffolding sequences.

Community
Advanced

Frequently Asked Questions

FAQPage Schema
How to install claude-education-skills?β–Ό

Run `npx skills add GarethManning/claude-education-skills --all -g -y` in your terminal to install the full library globally.

What can these education skills do for teachers?β–Ό

They generate lesson plans, hinge questions, rubrics, discussion protocols, feedback scripts, and tutoring dialogues, all grounded in named education research like Black & Wiliam and Sweller.

Do I need coding skills to use these education skills?β–Ό

No. You can copy any skill's prompt block and paste it into any AI chat, then fill in your class details in plain English.

Which AI tools work with this skills library?β–Ό

The skills follow the Agent Skills standard and work in Claude Code, Claude.ai via MCP, OpenAI Codex, Hermes Agent, and any compatible tool.

Are the teaching methods in these skills research-based?β–Ό

Yes. Every skill cites named studies and rates its evidence strength as strong, moderate, emerging, or original, with poorly evidenced approaches deliberately excluded.

Related Repositories in Education & Research

View All in Education & Research→