The vocabulary of context-aware engineering
Plain-language definitions of the concepts behind code-anchored context, repo memory, and AI-native engineering.
Context drift
Context drift is when written documentation no longer matches the code it describes — the docs say one thing, the system does another.
Code-anchored context
Code-anchored context is team knowledge attached directly to the files and globs it describes, instead of living in a separate wiki.
Repo memory
Repo memory is durable team knowledge about a codebase — decisions, gotchas, invariants — that persists across sessions, people, and AI agents.
AGENTS.md
AGENTS.md is a convention for a repository file that tells AI coding agents how to work in that codebase — the agent-readable counterpart to a README.
Context integrity
Context integrity is the property that a team’s recorded knowledge about a codebase still matches what the code actually does — and stays that way as the code changes.
Context engineering
Context engineering is the practice of deliberately assembling the right information for a human or AI agent before it acts — deciding what context to deliver, when, and in what form.
Living documentation
Living documentation is documentation that stays in sync with the system it describes, because it is connected to that system and re-checked when the system changes — rather than written once and left to rot.
Model Context Protocol (MCP)
The Model Context Protocol (MCP) is an open standard for connecting AI applications to external tools and data sources through a single uniform interface.
Architecture Decision Record (ADR)
An Architecture Decision Record (ADR) is a short document that captures one significant technical decision — its context, the choice made, and its consequences — so the reasoning survives the decision.
Tribal knowledge
Tribal knowledge is the undocumented understanding of how a system really works that lives only in the heads of experienced team members — invisible, unsearchable, and lost when they leave.
Context coverage
Context coverage is the share of a codebase that has documented, code-anchored team knowledge attached to it — how much of the system a new engineer or agent can be briefed on before touching it.
Generative engine optimization (GEO)
Generative engine optimization (GEO) is the practice of making your content more likely to be cited in answers from AI systems like ChatGPT, Claude, and Perplexity — the AI-era counterpart to SEO.
Prompt engineering
Prompt engineering is the practice of crafting the instructions given to an AI model to get reliable, high-quality output — tuning the request itself, as distinct from the context the model works on.
Retrieval-augmented generation (RAG)
Retrieval-augmented generation (RAG) is a technique that fetches relevant documents at query time and feeds them to a language model, so its answer is grounded in specific sources rather than only its training data.
Answer engine optimization (AEO)
Answer engine optimization (AEO) is the practice of structuring content so AI answer engines can extract a direct, citable answer from it — closely related to GEO, with the emphasis on being the answer rather than a link.
Vector database
A vector database stores and searches high-dimensional embeddings by similarity, powering semantic search and retrieval-augmented generation for AI applications.
Semantic search
Semantic search ranks results by meaning rather than exact keyword match, using embeddings so a query and a document can match even when they share no words.