Visit the interactive documentation at: https://dmccreary.github.io/ibook-skills/
Agent Skills for Intelligent Textbooks is a portable, token-efficient library of AI agent skills for building interactive educational textbooks. The skills are written as plain-language markdown workflows (SKILL.md files), so they aren't locked to one vendor's assistant — the same skill set has been installed and run in Claude Code, OpenAI Codex, Google Antigravity/Gemini, Cursor, Perplexity, and Hermes. The repository's CLAUDE.md instructions file doubles as an AGENTS.md for tools that use that convention, and dedicated install scripts (bk-install-skills, bk-install-skills-codex, bk-install-skills-antigravity) symlink the same skills into each tool's own skills directory.
This project enables the creation of Level 2+ intelligent textbooks using MkDocs with the Material theme, incorporating learning graphs, concept dependency mapping, interactive p5.js simulations (MicroSims), and AI-assisted content generation. Every skill follows educational best practices including Bloom's Taxonomy (2001 revision) for learning outcomes, ISO 11179 standards for terminology definitions, and concept dependency graphs (DAGs) to ensure prerequisites are taught before they're used.
Whether you're an educator building course materials, a technical writer producing documentation, or a developer exploring educational technology, these agent skills provide a systematic, repeatable pipeline from a course description all the way to a published, interactive textbook — regardless of which AI coding agent you point at the repository.
| Metric | Count |
|---|---|
| Concepts in Learning Graph | 570 |
| Chapters | 31 |
| Appendices | 3 |
| MicroSims | 94 |
| Glossary Terms | 570 |
| FAQ Questions | 66 |
| Diagrams | 25 |
| Equations | 131 |
| Total Words | 154,857 |
| Equivalent Pages | ~672 |
Chapter-level quizzes and curated references are actively being generated and are not yet complete across all 31 chapters.
Skills Available: 14 active skills (several are meta-skills that route to dozens of sub-workflows) covering learning graph generation, chapter content, MicroSims, glossaries, FAQs, quizzes, references, media, and publishing.
- Python 3.8 or higher
- Git
- An AI coding agent that supports agent skills or a similar convention (e.g. Claude Code, OpenAI Codex, Google Antigravity, Cursor)
- Basic familiarity with markdown and the command line
git clone https://github.com/dmccreary/ibook-skills.git
cd ibook-skillsThis project uses MkDocs with the Material theme:
pip install mkdocs
pip install mkdocs-material
pip install pymdown-extensionsBuild the documentation site:
mkdocs buildServe locally for development (with live reload):
mkdocs serveOpen your browser to http://localhost:8000
mkdocs gh-deployThis will build the site and push it to the gh-pages branch.
Set BK_HOME to the repository root, then run the installer for your platform:
export BK_HOME="$HOME/path/to/ibook-skills"
# Claude Code
$BK_HOME/scripts/bk-install-skills
# OpenAI Codex
$BK_HOME/scripts/bk-install-skills-codex
# Google Antigravity / Gemini
$BK_HOME/scripts/bk-install-skills-antigravityEach script symlinks every active skill in skills/ (skipping skills/archived/) into that tool's own skills directory, so edits to a skill here are picked up immediately by every connected agent.
List available skills:
$BK_HOME/scripts/bk-list-skillsInvoking a skill: most agents pick up a skill automatically once the task matches its description. Where a tool supports explicit invocation, reference it by name, e.g. Use the learning-graph-generator skill to create a concept graph for my course.
Navigation:
- Use the left sidebar to browse chapters and skill descriptions
- Click the search icon (🔍) to search all content
- The Learning Graph section shows concept dependencies and quality metrics
Interactive MicroSims:
- Found throughout the chapters and in the "MicroSims" section of the documentation
- Each simulation runs standalone in your browser
- Adjust parameters with sliders and interactive controls
- View source code and customize for your own use
Customization:
- Edit markdown files in
docs/to modify content - Modify
mkdocs.ymlto change site structure and navigation - Add your own MicroSims in
docs/sims/[microsim-name]/ - Customize theme colors and styles in
docs/css/extra.css - Create custom skills in
skills/[skill-name]/SKILL.md
ibook-skills/
├── docs/ # MkDocs documentation source
│ ├── chapters/ # 31 chapters + appendices on building intelligent textbooks
│ │ ├── 01-foundations-ai-language-models/
│ │ │ └── index.md # Chapter content
│ │ └── ...
│ ├── sims/ # 94 interactive MicroSims (p5.js, Chart.js, vis-network, ...)
│ ├── learning-graph/ # Learning graph data, analysis, and book metrics
│ │ ├── concept-list.md # 570 concepts enumerated
│ │ ├── learning-graph.csv # Concept dependencies (DAG)
│ │ ├── learning-graph.json # vis-network JSON format
│ │ ├── book-metrics.json # Canonical book-wide metrics (source of truth)
│ │ └── quality-metrics.md # Graph quality analysis
│ ├── glossary.md # ISO 11179-compliant definitions (570 terms)
│ └── faq.md # Frequently asked questions (66 Q&A)
├── skills/ # AI agent skill definitions
│ ├── learning-graph-generator/ # Generates 300-600 concept dependency graphs
│ ├── book-chapter-generator/ # Designs chapter structure from the learning graph
│ ├── chapter-content-generator/ # Generates chapter content and MicroSims
│ ├── glossary-generator/ # ISO 11179-compliant glossary
│ ├── faq-generator/ # FAQs from chapter content
│ ├── quiz-generator/ # Bloom's Taxonomy-aligned quizzes
│ ├── reference-generator/ # Curated per-chapter references
│ ├── microsim-generator/ # MicroSims across many JS libraries (meta-skill)
│ ├── microsim-utils/ # MicroSim QA and maintenance (meta-skill)
│ ├── book-installer/ # Project scaffold and infrastructure (meta-skill)
│ ├── book-media-generator/ # Slides, illustrations, audio (meta-skill)
│ ├── book-publisher/ # README, LinkedIn, press release (meta-skill)
│ └── archived/ # Original single-purpose skills, consolidated into the meta-skills above
├── scripts/ # Utility scripts (install, metrics, screenshots, ...)
│ ├── bk-install-skills # Install skills for Claude Code
│ ├── bk-install-skills-codex # Install skills for OpenAI Codex
│ ├── bk-install-skills-antigravity # Install skills for Google Antigravity/Gemini
│ └── bk-list-skills # List all active skills
├── commands/ # Slash commands
├── mkdocs.yml # MkDocs configuration
├── CLAUDE.md # Agent instructions (also usable as AGENTS.md)
└── README.md # This file
Found a bug, typo, or have a suggestion for improvement? Please report it:
When reporting issues, please include:
- Clear description of the problem or suggestion
- Steps to reproduce (for bugs)
- Expected vs actual behavior
- Screenshots or error messages (if applicable)
- Skill name and the AI agent/tool you were using (if skill-specific)
- Browser/environment details (for MicroSim issues)
This work is licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
You are free to:
- Share — Copy and redistribute the material in any medium or format
- Adapt — Remix, transform, and build upon the material
Under the following terms:
- Attribution — Give appropriate credit with a link to the original repository
- NonCommercial — No commercial use without explicit permission
- ShareAlike — Distribute your contributions under the same license
Attribution Example:
This work is based on "Agent Skills for Intelligent Textbooks" by Dan McCreary,
available at https://github.com/dmccreary/ibook-skills, licensed under CC BY-NC-SA 4.0.
See license details for the full legal text.
This project is built on the shoulders of giants in the open source community. We are deeply grateful to:
- MkDocs - Fast, simple static site generator optimized for project documentation
- Material for MkDocs - Beautiful, responsive Material Design theme with advanced features
- p5.js - Creative coding library from NYU ITP for interactive educational simulations
- vis-network - Network visualization library for learning graph exploration
- Chart.js - Interactive charts for data visualization
- Mermaid - Diagram and flowchart generation from text
- Python community - Data processing, analysis, and automation tools
- PyMdown Extensions - Enhanced Markdown extensions
- GitHub Pages - Free hosting for open source documentation projects
- GitHub - Version control and collaboration platform
- Bloom's Taxonomy - Framework for categorizing cognitive learning objectives
- ISO 11179 - International standard for metadata registries (glossary definitions)
- Dublin Core - Metadata standards for educational resources
Special thanks to the educators, developers, and maintainers who contribute to making educational resources accessible, interactive, and open to all. Your work enables projects like this to exist and thrive.
Dan McCreary
- LinkedIn: linkedin.com/in/danmccreary
- GitHub: @dmccreary
Questions, suggestions, or collaboration opportunities? Feel free to connect on LinkedIn or open an issue on GitHub. I'm particularly interested in:
- Feedback on skill effectiveness and quality across different AI agents
- Suggestions for new skills or features
- Collaboration on educational technology projects
- Use cases and success stories from educators
If you use these agent skills in your research, teaching, or projects, please cite:
McCreary, D. (2024). Agent Skills for Intelligent Textbooks. GitHub.
https://github.com/dmccreary/ibook-skills
BibTeX:
@misc{mccreary2024agentskills,
author = {McCreary, Dan},
title = {Agent Skills for Intelligent Textbooks},
year = {2024},
publisher = {GitHub},
url = {https://github.com/dmccreary/ibook-skills},
note = {A portable, multi-platform library of AI agent skills for creating intelligent educational content}
}This repository includes 14 active skills for intelligent textbook creation. Several are meta-skills that route to focused sub-workflows in a references/ folder, which keeps the total under each platform's loaded-skills limit while still covering dozens of tasks.
- course-description-analyzer - Validates and enhances course descriptions
- learning-graph-generator - Creates 300-600 concept dependency graphs (DAG structure)
- book-chapter-generator - Designs chapter structure from the learning graph
- chapter-content-generator - Generates chapter content, diagrams, and exercises
- glossary-generator - Creates ISO 11179-compliant glossaries
- faq-generator - Generates FAQs from chapter content and the learning graph
- quiz-generator - Creates Bloom's Taxonomy-aligned quiz questions
- reference-generator - Curates per-chapter academic references
- book-installer - Project scaffolding and infrastructure: init a new textbook, install features (math, mascot, learning-graph viewer, Google Analytics, and more), generate book metrics
- microsim-generator - MicroSims across p5.js, Chart.js, Plotly, Mermaid, vis-network, timelines, maps, and more
- microsim-utils - MicroSim QA: standardization, screenshots, layout review, coverage reports
- book-media-generator - Slide decks, illustrated stories, chapter images, and audio
- book-publisher - README, LinkedIn posts/carousels, and press releases
- docx-to-web-publisher - Converts .docx files into structured web pages
For detailed documentation on each skill, see the skill descriptions in the documentation site.
Portable Agent Skills | Works with Claude Code, Codex, Antigravity/Gemini, Cursor, Perplexity, and Hermes | Open Educational Resources