Activity Index

Every activity deck in this course, 106 of them, in one place.

The course schedule now lists at most three optional links per session, so that the required reading for any given day is unmistakable. Everything else lives here. Nothing on this page is required unless a session or an assignment says so — this is a library, not a backlog.

How to use it. If a lab direction or a class session names a deck, follow that link. Otherwise browse by theme when you want to go deeper on something that caught your attention, or when you are choosing a project direction and want to know what is possible.

Agents and orchestration

Activity	Length	Role
AI Coding Agent Security: Poisoned Repos, the Software Supply Chain, and State-of-the-Art Defenses	~4,063 words	reference
Advanced Agent Loops: Control Flow, Reflection, and Recovery	~4,583 words	on the schedule
Agent Frameworks: LangChain, CrewAI, AutoGen, Agno, and DeepAgents	~7,735 words	on the schedule
Agent Security: Threat Modeling and the OWASP LLM Top 10	~5,520 words	on the schedule
Agent Skills and Plugins: Building, Configuring, and Publishing Custom Capabilities	~6,787 words	on the schedule
Agent Teams: Specialists over Monoliths	~3,118 words	reference
Agentic CLI Tools: Claude Code, Codex, Gemini, opencode, pi, and Friends	~7,737 words	on the schedule
Agentic Case Studies: Migration, Browsing, and Research Agents	~4,032 words	on the schedule
Coding Agents: Agentic Development Tools	~6,694 words	on the schedule
Connecting Agents to the World: MCP and APIs	~4,305 words	on the schedule
Connecting Agents to the World: MCP, REST APIs, and OAuth 2.0	~4,518 words	on the schedule
Debugging AI Agents: What Went Wrong and How to Find Out	~5,389 words	reference
Designing Agent Personas and System Prompts	~3,238 words	on the schedule
From Second Brain to Chief of Staff: A Personal Agent in Production	~5,186 words	on the schedule
GitHub Superpowers for AI Developers	~4,066 words	on the schedule
Governing Coding Agents: Charters, Handoffs, and Durable Memory	~5,167 words	reference
Human-in-the-Loop: Oversight, Escalation, and Appropriate Autonomy	~4,468 words	on the schedule
Memory and the Small Context Window Principle	~4,271 words	on the schedule
Memory in Agents: What They Remember and Why It Matters	~4,513 words	on the schedule
Multi-Agent Communication: Protocols, Shared State, and Coordination	~4,532 words	on the schedule
Multi-Agent Debate	~3,552 words	on the schedule
Multimodal Agents: Vision, Documents, and Code as First-Class Inputs	~5,516 words	reference
Orchestration Patterns: Pipelines, Routers, and Planners	~6,679 words	on the schedule
Red-Teaming LLMs: Finding Failures Before Deployment	~4,941 words	reference
Scheduled and Autonomous Agent Runs: Cron, systemd, n8n, and GitHub Actions	~3,103 words	reference
Stochastic Multi-Agent Consensus	~2,651 words	on the schedule
Supplemental Tutorial: Agentic OpenWebUI — Tools, Uploads, and Multi-Agent Workflows	~3,523 words	on the schedule
Testing Agents: Evaluation, Regression, and the Non-Determinism Problem	~3,665 words	on the schedule
The Agent Loop: Perceive, Plan, Act	~3,421 words	on the schedule
The Critique and Refine Pattern	~3,728 words	on the schedule
The Economics of AI Agents: Pricing, Incentives, and Market Dynamics	~5,191 words	reference
The Local Agent Stack: Wiring Containers into a System	~4,475 words	on the schedule
Tool Use and Function Calling	~6,094 words	on the schedule
Visual Agent Building with Langflow	~3,641 words	on the schedule

How models work

Activity	Length	Role
Anatomy of an LLM Request: One Prompt, End to End, by Hand	~4,030 words	on the schedule
Attention and Transformers, Conceptually and by Hand	~3,600 words	on the schedule
Evaluating Agents: LLM-as-Judge and Rubric Pipelines	~3,103 words	on the schedule
From Rewards to Preferences: Reinforcement Learning and RLHF	~6,468 words	reference
From Text Generation to a Neural Network	~4,695 words	on the schedule
How LLMs Are Built: Tokenization, Pre-Training, and Scaling	~6,651 words	on the schedule
Model Types and Model Lifecycle: What “Thinking,” “Reasoning,” and “Vision” Actually Mean	~3,884 words	reference
Privacy-Preserving AI: Federated Learning, Differential Privacy, and PII Scrubbing	~5,206 words	on the schedule
RESTful LLM Access: The api/v1 Paradigm	~7,377 words	on the schedule
Serving LLMs in Production: Latency, Cost, and Throughput	~6,780 words	reference
Shakespeare GPT: How Language Models Learn to Write	~3,525 words	on the schedule
Temperature and Sampling Explorer: Hands-On Parameter Tuning	~3,329 words	on the schedule
The Hardware Behind AI: GPUs, Quantization, and Running Models at the Edge	~5,749 words	on the schedule
The LLM Wiki: Karpathy’s Pattern and a Vault Full of Use Cases	~5,553 words	on the schedule
Tokens and Embeddings: How Agents Represent Meaning	~2,649 words	on the schedule
Traditional Machine Learning Foundations	~2,688 words	on the schedule
Why Different Answers Every Time? Sampling, Temperature, and Generation	~3,129 words	on the schedule

Retrieval and knowledge

Activity	Length	Role
Fine-Tuning, RAG, and Prompting: Choosing the Right Approach	~3,877 words	on the schedule
RAG Quality: Chunking, Clustering, and Reranking	~4,872 words	on the schedule
Retrieval-Augmented Generation with Chroma	~2,935 words	on the schedule
Syncing Obsidian to GitHub and Wiring AI Agents to Your Vault	~6,665 words	on the schedule
Synthetic Data: Using AI to Train AI	~4,999 words	on the schedule
The Second Brain: Obsidian, Gitless GitHub Sync, and Agent Access	~5,229 words	on the schedule
Vector Databases: How Agents Search for Meaning	~4,163 words	on the schedule

Evaluation and reliability

Activity	Length	Role
Agent Observability and Tracing	~3,366 words	reference
Benchmark Design: How We Know If AI Systems Work	~4,386 words	reference
Explainability and Human-Centric Design	~3,010 words	reference
Explainability in Depth: SHAP, LIME, Attention, and the Limits of Interpretation	~5,346 words	reference
Hallucinations and Evaluating Agent Outputs	~2,879 words	on the schedule
Structured Outputs: JSON Mode, Tool Schemas, and Output Validation	~4,915 words	on the schedule
Why Doesn’t AI Give the Same Answer Twice? Deterministic and Probabilistic Computing	~2,793 words	on the schedule

Safety, security, and deployment

Activity	Length	Role
Adversarial Robustness and Red-Teaming LLMs	~4,324 words	on the schedule
Containerizing AI Systems: Safety, Isolation, and Trust Boundaries	~4,521 words	on the schedule
Cost Optimization: Token Budgets, Caching, and Model Routing	~4,138 words	on the schedule
Deploying Agents: From Laptop to Production	~5,311 words	reference
Designing Your AI Development Environment	~3,925 words	on the schedule
Docker from Zero: Containers for Agent Builders	~4,983 words	on the schedule
Hosting with Cloudflare: Workers, Pages, and Wrangler	~9,429 words	on the schedule
Prompt Injection and Agent Security: The OWASP LLM Top 10	~4,656 words	on the schedule
Running Your Own AI: Ollama, OpenWebUI, and Private Local Models	~3,152 words	on the schedule
Terminal and Filesystem Isolation for Agent Safety	~4,160 words	on the schedule
The Course Development Environment: One Container for Every Lab	~2,980 words	on the schedule
The Local Model Landscape: Llama, Mistral, Phi, Gemma, and Friends	~4,652 words	on the schedule
The Shell: Your Agent’s Native Habitat	~5,293 words	on the schedule

Ethics, policy, and society

Activity	Length	Role
AI Alignment and Safety: From RLHF to Constitutional AI	~3,424 words	on the schedule
AI Creativity: Generative Models, Authorship, and the Nature of Originality	~5,177 words	on the schedule
AI Regulation: EU AI Act, NIST AI RMF, and Sector Rules	~4,492 words	reference
AI and the Future of Work: Automation, Displacement, and Adaptation	~4,085 words	reference
AI for Accessibility: Opportunity, Obligation, and Risk	~3,970 words	on the schedule
AI in Education: Opportunity, Disruption, and the Integrity Question	~4,250 words	reference
Case Study: Early-Alert and At-Risk Identification with Local AI	~4,815 words	reference
Dataset Documentation: Model Cards and Datasheets	~4,773 words	reference
Environmental Impact and the Carbon Cost of Intelligence	~5,209 words	on the schedule
Ethical Frameworks for Agentic AI Systems	~5,552 words	reference
Governance and Policy Writing	~2,762 words	on the schedule
Integrating AI with Institutional Systems: LMS, Assessment Software, and Student Information Systems	~4,321 words	reference
Intellectual Property, Privacy, and the Case for Local AI	~2,661 words	on the schedule
Neuro-AI Ethics: Brain-Inspired AI, Cognitive Science, and the Study of Mind	~4,603 words	on the schedule
The Philosophy and Psychology of Artificial Intelligence	~4,149 words	reference
Training Data and Bias	~3,890 words	on the schedule

Building and shipping

Activity	Length	Role
AI for Scientific Discovery: AlphaFold, Drug Discovery, and the Reproducibility Crisis	~4,272 words	reference
AI-Assisted Development: The “Let Me Cook” Approach	~4,148 words	on the schedule
Building an AI Chess Coach: LLM API Calls in a Real Web App	~6,325 words	reference
Design First: Plan Before You Build	~4,397 words	on the schedule
Project Studio and Gallery Walk	~2,985 words	on the schedule
Publishing Your Work: GHCR, Docker Hub, and npm	~5,246 words	reference
Structured Peer Review: SQR Cards and Team Check-Ins	~3,231 words	on the schedule
Supplemental Tutorial: A Threaded Auto-Research Pipeline	~3,451 words	reference
Team Charter and Norms	~3,160 words	on the schedule
The AI Maker: Agents, Engineering Discipline, and Human-Centric Design	~5,540 words	reference
Using AI for Academic Research	~2,676 words	reference

Other

Activity	Length	Role
Prompt Engineering as Agent Design: Personas and System Prompts	~3,886 words	on the schedule