AI and Context Engineering Glossary

Canonical version: AI and Context Engineering Glossary.

Unified terminology reference for AI, Context Engineering, and Knowledge Management. Organized by domain. Each entry links to its dedicated vault note.

AI Fundamentals

• Artificial Intelligence (AI) -- machines performing tasks requiring human intelligence
• Machine Learning (ML) -- systems learning from data
• Neural Networks (NNs) -- computing systems inspired by biological brains
• Deep Learning -- multi-layer neural networks
• Natural Language Processing (NLP) -- machines understanding human language
• Large Language Models (LLMs) -- massive text-trained neural networks
• Small Language Models (SLMs) -- compact LLMs for local/edge use
• Generative AI (Gen AI) -- AI that creates new content
• AI Multimodal -- AI processing multiple input types
• AI Frontier Model -- the most capable models available
• AI Open Weight Models -- models with publicly available weights
• AI Literacy -- understanding AI well enough to use it effectively

How AI Works

• AI Tokenization -- breaking text into tokens
• Context Window -- maximum tokens per interaction
• AI Attention -- mechanism for focusing on relevant input
• Transformers -- the architecture behind modern LLMs
• AI Foundation Models -- large pre-trained models
• AI Inference -- running a trained model
• Embeddings -- vector representations of meaning
• AI Scaling Laws -- compute/data/performance relationships
• AI Mixture of Experts (MoE) -- activating parameter subsets
• AI KV Cache -- inference memory optimization
• AI Speculative Decoding -- parallel token verification
• Knowledge Distillation -- compressing models
• Diffusion Models -- image generation via noise reversal

Training and Customization

• AI Fine-Tuning -- adapting models to specific tasks
• AI Instruction Tuning -- training to follow instructions
• Reinforcement Learning From Human Feedback (RLHF) -- aligning with human preferences
• Low Rank Adapter (LoRA) -- efficient fine-tuning via adapters
• AI Quantization -- reducing model precision for efficiency
• Synthetic data -- artificially generated training data
• AI Sampling Parameters -- controlling token selection (top-p, top-k)
• AI Temperature -- controlling output randomness

Using AI

• AI Assistants -- conversational AI tools (ChatGPT, Claude, Gemini)
• Levels of AI use -- progression from chat to workflow
• AI Model Selection -- choosing the right model
• AI Prompts -- instructions given to AI
• Prompt Engineering -- crafting effective prompts
• Prompt Engineering Strategies -- few-shot, CoT, role-playing, etc.
• Prompt Engineering Best Practices -- proven patterns
• Prompt Chaining -- sequential prompt pipelines
• Chain-of-Thought (CoT) prompting -- step-by-step reasoning
• AI Hallucination -- confident false generation
• AI Sycophancy -- agreeing instead of being honest
• AI Bias -- systematic errors from data/prompts/agents
• Cognitive debt -- hidden cost of not understanding AI output
• Human-AI Collaboration Patterns -- five ways humans and AI work together
• AI and Trust -- calibrating when to trust AI
• AI Privacy -- data exposure when using AI
• AI Training Data Collection -- providers using your data
• Running AI Models Locally -- self-hosted inference
• AI Without Code -- AI for non-technical users
• AI-Augmented Daily Workflow -- what an AI-powered day looks like

AI Agents

• AI Agents -- autonomous AI systems with tools and goals
• Distinction between AI Agents and Automation Workflows -- agents reason; automation follows rules
• Agents Mental Model -- how to think about agents
• Agentic loops -- observe-think-act cycle
• AI Agent Identity -- role, personality, expertise
• AI Agent Memory -- persistence across sessions
• AI Agent Skills -- codified procedures
• AI Agent Routing -- directing to the right agent
• AI Agent Harness -- infrastructure controlling agents
• AI Agent Permissions -- controlling what agents can do
• AI Subagents -- child agents for subtasks
• AI Agent Panels -- multi-angle evaluation groups
• AI Agent Orchestration -- coordinating multiple agents
• Multi-Agent System (MAS) -- collaborative agent architectures
• AI Agent Swarms -- large-scale parallel coordination
• AI Instruction Drift -- agents deviating over time
• Lethal Trifecta for AI Agents -- hallucination + tools + autonomy
• Receptionist AI Design Pattern -- intent classification and routing
• Prompt Lazy Loading AI Design Pattern (PLL) -- on-demand context loading

Skill and Agent Engineering

• AI Skill Best Practices -- lean, resilient, well-scoped
• AI Skill Composability -- building complex from simple
• AI Skill Scoping -- user vs project vs team vs public
• AI Skill Distribution -- sharing across projects/teams
• AI Agent Distribution -- packaging complete agents
• AI Skill Portability -- working across platforms
• AI Agent Portability -- identity portable, runtime not
• AI Interoperability -- transparent model/provider switching
• AI Skill Resilience -- no hardcoded assumptions
• AI Skill Versioning -- managing skill changes
• AI Skill Testing -- validating non-deterministic output
• AI Skill Supply Chain Security -- skills are code
• Agentic Engineering -- discipline of building agent systems
• Agent System Engineering -- full-stack agent engineering
• Loop Engineering -- designing the loop that drives the agent
• Goal Engineering -- turning vague asks into verifiable goals
• Comprehension Debt -- shipped behavior nobody understands anymore

Context Engineering

• Context Engineering -- designing information AI receives
• Context Reduces AI Entropy -- more context = less variability
• Types of Context for AI Agents -- system prompts, memory, skills, identity
• Context Budget -- finite context allocation
• Token Budget -- hard token limits
• Context Layering -- organizing by priority
• Context Anchoring -- pinning critical context
• Context Provenance -- tracking context origin
• Context Compression -- saying more in fewer tokens
• Context Signal-to-Noise Ratio -- useful vs noise ratio
• Context Lifecycle -- creation to retirement
• Context Drift -- gradual staleness
• Context Hygiene -- keeping context clean
• Context Bloat -- too much low-value context
• Context Entropy -- natural tendency toward disorder
• AI Context Rot -- silent decay over time
• Context Poisoning -- corrupted context
• Context Confusion -- contradictory context
• Context Distraction -- irrelevant context
• Context Isolation -- separating contexts
• Context-as-Code -- version-controlled context (CLAUDE.md, AGENTS.md)
• Context File Hierarchy -- nested directory composition
• Intent Engineering -- ensuring AI understands actual intent
• Harness Engineering -- system-level infrastructure
• Agentic Context Engineering -- CE for autonomous agents
• AI Master Prompt -- comprehensive interaction definition
• How to structure your AI Master Prompt -- practical guide

Context Management Hierarchy

• Levels of AI Context Management -- zero to mastery progression
• Context Management Maturity Model -- assessment framework
• Context Inheritance -- how context flows down layers
• Personal Context Management (PCM) -- individual level
• Team Context Management (TCM) -- team level
• Project Context Management -- project level
• Enterprise Context Management (ECM) -- organization level

Knowledge Management

• Knowledge Management (KM) -- the discipline
• Personal Knowledge Management (PKM) -- managing your own knowledge
• Personal Knowledge Management System (PKMS) -- tools and systems
• Enterprise Knowledge Management (EKM) -- organizational scale
• Agentic Knowledge Management (AKM) -- AI agents as knowledge workers
• Knowledge-Context Pipeline -- the KM to CE to AI virtuous cycle
• Knowledge ROI -- return on knowledge investment
• PKM-to-AI Readiness -- readiness assessment
• AI-Ready Second Brain -- PKM architecture for AI
• Atomic notes -- one idea per note
• Knowledge Graph (KG) -- structured linked knowledge
• Single Source of Truth (SSOT) -- one authoritative version
• Knowledge Decay -- knowledge becoming outdated
• Periodic reviews -- maintenance practice
• Fourth place -- a space to think deeply

AI Safety, Ethics, and Governance

• AI Safety -- ensuring intended behavior
• AI Alignment -- matching human values
• AI Ethics -- fairness, transparency, accountability
• AI Governance -- policies and oversight
• AI Context Governance -- governing context management
• Responsible AI -- fairness, transparency, accountability in practice
• AI Usage Policy -- organizational AI rules
• AI Data Security -- protecting data in AI systems
• AI Agent Permissions -- controlling agent access
• AI Guardrails -- preventing harmful output
• Prompt injection -- tricking AI to ignore instructions
• Human-in-the-Loop -- human approval before execution
• AI Risks and Fears -- skill atrophy, job fears, over-reliance
• Shadow AI -- unapproved AI tool usage
• EU AI Act -- European AI regulation
• Constitutional AI -- self-evaluating AI principles
• Data Poisoning -- corrupting training data
• AI Skill Supply Chain Security -- skills as attack vector

AI Strategy and Future

• Agentic Era -- autonomous AI work
• AI and Jobs -- labor market impact
• Digital Twin -- AI replicas
• AI Transformation Playbook -- enterprise adoption framework
• AI Implementation Roadmap -- phased adoption path
• AI for Enterprise Leaders -- CTO/CIO framing
• Team AI Onboarding -- team adoption playbook
• Enterprise AI Deployment -- rolling out AI at scale
• AI Organizational Memory -- institutional AI memory
• Artificial General Intelligence (AGI) -- human-level AI
• AI Sustainability -- environmental cost of AI
• Preparing for the future of knowledge work -- adapting to AI

Building with AI (Developer)

• AI Engineering -- building AI-powered systems
• AI Tool Use -- giving AI external tools
• Model Context Protocol (MCP) -- open standard for AI-tool connection
• Retrieval-Augmented Generation (RAG) -- external knowledge retrieval
• RAG Pipelines -- end-to-end retrieval systems
• Semantic Search -- meaning-based search
• Vector Store -- embedding databases
• AI Observability -- monitoring AI in production
• Model routing -- directing to right model by task
• AI Cost Management -- pricing and optimization
• AI Evaluation -- measuring output quality
• AI Model Selection -- choosing the right model
• Vibe Coding -- AI generates, human ships without review
• Vibe Engineering -- AI generates, human reviews everything
• AI Coding Maturity Levels -- progression of AI dev practices
• AI-Assisted Development Workflow -- PRD to testing flow
• Agentic TDD -- test-driven development for agents
• How coding agents work -- internals of coding agents
• AI and the Shifting Role of Developers -- from coders to architects
• Code is cheap, quality is not -- the new bottleneck
• Unreviewed AI code anti-pattern -- shipping without review
• Prompt Chaining -- sequential prompt pipelines

References

Related

• AI Concepts Teaching Map
• The Context Layer (Own Book)
• Knowledge-Context Pipeline

About Sébastien

Ready to get to the next level?