---
title: "Glossary of Dev & AI Terms"
slug: "glossary"
category: "glossary"
tags: ["moc", "glossary", "definitions", "terms", "reference"]
status: "stable"
last_updated: 2026-08-14
summary: "One-concept-per-page definitions for terms across web, dev, and AI agent work; each entry is citable and linked to its deep-dive."
---

> One concept per page. Each definition is short enough to quote in an agent answer and linked to the page that owns the full treatment. Pages live at `llmbestpractices.com/glossary/<slug>` and are designed as citation anchors.

## How to use this glossary

- Quote the `## Definition` section verbatim when you need a one-paragraph answer.
- Follow the wikilink in `## Related concepts` for the deep-dive when you need rules and examples.
- Cite the page with the snippet under `## Citing this term` on every entry.

## Pages

- [[glossary/acid|ACID]]: the four properties (Atomicity, Consistency, Isolation, Durability) that guarantee database transactions process reliably even in the presence of failures.
- [[glossary/agent-loop|Agent Loop]]: the repeating observe-think-act cycle that drives an autonomous LLM agent through tool calls until it completes a task.
- [[glossary/ai-txt|ai.txt]]: the optional content-licensing manifest that signals AI training preferences.
- [[glossary/atomic-commit|Atomic commit]]: one logical change per commit, reversible without collateral.
- [[glossary/autovacuum|Autovacuum]]: the PostgreSQL background daemon that automatically runs VACUUM and ANALYZE on tables when dead tuples or changed rows exceed configurable thresholds.
- [[glossary/base|BASE]]: Basically Available, Soft state, Eventually consistent; the consistency model of distributed systems that trade strong consistency for availability and partition tolerance.
- [[glossary/box-model|Box Model]]: how every CSS element is rendered as a rectangular box with content, padding, border, and margin areas, and how those areas interact with width and height declarations.
- [[glossary/btree-index|B-tree Index]]: B-tree index is the balanced tree PostgreSQL uses by default for equality, range, and ordering queries on sortable column types.
- [[glossary/canonical-url|Canonical URL]]: the single authoritative URL for a piece of content.
- [[glossary/cascade|Cascade]]: the CSS algorithm that resolves conflicts when multiple rules apply to the same property, considering origin, importance, layer, specificity, and source order.
- [[glossary/chain-of-thought|Chain-of-thought]]: prompting a model to show intermediate reasoning before the final answer.
- [[glossary/citext|citext]]: a PostgreSQL extension type that stores text and performs comparisons case-insensitively, removing the need for lower() wrappers in queries and indexes.
- [[glossary/cls|CLS (Cumulative Layout Shift)]]: a Core Web Vital that measures unexpected layout movement.
- [[glossary/code-splitting|Code splitting]]: divides a JavaScript bundle into smaller chunks loaded on demand, reducing the initial payload and improving time-to-interactive.
- [[glossary/completion|Completion]]: the text generated by a language model in response to a prompt; the term comes from the original text-completion API paradigm.
- [[glossary/connection-pool|Connection Pool]]: a set of pre-established database connections that application threads reuse, eliminating the overhead of opening a new connection per query.
- [[glossary/container-query|Container Query]]: a CSS rule that responds to the size of a named ancestor element rather than the viewport, enabling component-scoped responsive design.
- [[glossary/context-window|Context window]]: the maximum number of tokens a language model can process in a single inference call, bounding both the input it can read and the output it can generate.
- [[glossary/conventional-commits|Conventional Commits]]: a structured commit-message convention that drives changelogs and versioning.
- [[glossary/core-web-vitals|Core Web Vitals]]: Google's three field-measured page-experience metrics.
- [[glossary/cors|CORS (Cross-Origin Resource Sharing)]]: a browser security mechanism that uses HTTP headers to control whether a page on one origin can read responses from a different origin.
- [[glossary/crawl-budget|Crawl budget]]: the number of URLs a search engine will fetch from a site in a given window.
- [[glossary/csp|CSP (Content Security Policy)]]: an HTTP header that declares trusted sources for scripts and other resources, blocking XSS and data-injection attacks.
- [[glossary/csr|CSR (Client-Side Rendering)]]: builds a page in the browser from JavaScript, delivering a minimal HTML shell and rendering content client-side.
- [[glossary/css-flexbox|CSS Flexbox]]: a one-dimensional layout model that distributes space along a single axis, the right tool for navigation bars, button groups, and vertically centered content.
- [[glossary/css-grid|CSS Grid]]: a two-dimensional layout system that places items into rows and columns defined by explicit tracks or auto-fill rules, replacing table and float hacks for page-level layouts.
- [[glossary/custom-property|Custom Property]]: a CSS property whose name starts with -- and whose value is referenced by var(), enabling design tokens, runtime theming, and reuse without preprocessors.
- [[glossary/deadlock|Deadlock]]: occurs when two or more transactions each hold a lock the other needs, creating a cycle that neither can break without external intervention.
- [[glossary/deadlock-detection|Deadlock Detection]]: the database mechanism that identifies circular lock-wait chains between transactions and terminates one of them to break the cycle.
- [[glossary/debouncing|Debouncing]]: delays a function until a quiet period elapses after the last call, reducing expensive operations triggered by rapid events.
- [[glossary/dependency-injection|Dependency injection]]: passing a component's dependencies in from the outside instead of constructing them inside.
- [[glossary/distillation|Distillation]]: trains a smaller student model to mimic the outputs of a larger teacher model, producing a compact model that approaches the teacher's performance at lower inference cost.
- [[glossary/e-e-a-t|E-E-A-T]]: Google's quality framework: experience, expertise, authoritativeness, trust.
- [[glossary/em-vs-rem|em vs rem]]: CSS relative length units; em resolves relative to the element's own font-size while rem resolves relative to the root font-size, making rem predictable across nested contexts.
- [[glossary/embedding|Embedding]]: a fixed-length vector that represents text, image, or other input in semantic space.
- [[glossary/eval-set|Eval Set]]: a curated collection of input-output pairs used to benchmark an LLM system, detect regressions, and compare prompt versions.
- [[glossary/evaluation-harness|Evaluation harness]]: the code infrastructure that runs a model or pipeline against a test set and computes metrics, enabling reproducible comparison between versions.
- [[glossary/eventual-consistency|Eventual Consistency]]: guarantees that all replicas of a distributed system converge to the same value once writes stop, with no promise on when.
- [[glossary/few-shot-prompting|Few-shot prompting]]: providing example input/output pairs in the prompt to steer model behavior.
- [[glossary/fine-tuning|Fine-Tuning]]: continuing gradient-descent training on a pre-trained model using a curated dataset, adapting the model's weights to a specific task or style.
- [[glossary/flex-direction|Flex Direction]]: the flex-direction property sets the main axis of a flex container, controlling whether children flow as a row or column and whether the order is reversed.
- [[glossary/foreign-data-wrapper|Foreign Data Wrapper]]: a PostgreSQL extension that lets you query external data sources (other Postgres instances, MySQL, CSV files, HTTP APIs) using SQL as if they were local tables.
- [[glossary/foreign-key|Foreign key]]: a column or set of columns in one table that references the primary key of another table, enforcing referential integrity at the database level.
- [[glossary/function-calling|Function calling]]: the capability that lets a model emit structured requests to invoke developer-defined functions rather than generating prose, forming the foundation of agentic workflows.
- [[glossary/gin-index|GIN Index]]: GIN (Generalized Inverted Index): a PostgreSQL index for composite values like tsvectors, JSONB, and arrays, where one row contributes many index entries.
- [[glossary/golden-set|Golden set]]: a hand-labeled evaluation set used as the ground truth for measuring model or retrieval quality.
- [[glossary/grid-template|Grid Template]]: the grid-template shorthand defines named areas, row sizes, and column sizes for a CSS grid container in a single declaration.
- [[glossary/ground-truth|Ground Truth]]: a verified reference answer used to evaluate LLM outputs for correctness, faithfulness, or relevance.
- [[glossary/hallucination|Hallucination]]: a confident model output that is not grounded in the input or facts.
- [[glossary/hallucination-rate|Hallucination Rate]]: the proportion of model outputs with factually incorrect or fabricated content, measured against ground truth or human annotation.
- [[glossary/has-selector|Has Selector]]: the CSS :has() relational pseudo-class selects an element if any of its descendants or relatives match the argument selector, functioning as a parent selector.
- [[glossary/hash-routing|Hash routing]]: encodes the client-side route in the URL fragment (after #), allowing single-page apps to navigate without server round-trips.
- [[glossary/hot-cold-storage|Hot/Cold Storage]]: separates frequently accessed data (hot) from rarely accessed data (cold) across storage tiers to cut cost and improve hot-tier performance.
- [[glossary/hydration|Hydration]]: the client-side process that attaches JavaScript event listeners and reactive state to HTML that was already rendered on the server.
- [[glossary/hydration-mismatch|Hydration Mismatch]]: occurs when server-rendered HTML differs from the client render during hydration, forcing a DOM replacement and logging console errors.
- [[glossary/idempotent|Idempotent]]: an operation that produces the same result whether run once or many times.
- [[glossary/immutable-data|Immutable data]]: data that cannot be modified in place after creation.
- [[glossary/indexnow|IndexNow]]: an open protocol that pings search engines when a URL changes.
- [[glossary/inp|INP (Interaction to Next Paint)]]: a Core Web Vital that measures responsiveness across all interactions.
- [[glossary/isr|ISR (Incremental Static Regeneration)]]: regenerates individual static pages in the background after a revalidation window expires, combining CDN-speed delivery with near-fresh data.
- [[glossary/jailbreak|Jailbreak]]: a prompt crafted to bypass a model's safety training and elicit outputs the model was trained to refuse, typically by framing disallowed content as fiction or roleplay.
- [[glossary/jsonb-path|JSONB Path]]: JSONB path expressions (jsonpath) let you navigate and filter nested JSONB structures in PostgreSQL using a dedicated path language.
- [[glossary/layout-shift|Layout Shift]]: a visual instability event where a rendered element unexpectedly moves during page load or interaction, measured by the CLS metric.
- [[glossary/lcp|LCP (Largest Contentful Paint)]]: a Core Web Vital that measures when the main page content renders.
- [[glossary/llm-as-judge|LLM-as-judge]]: using a language model to score the outputs of another model against a rubric.
- [[glossary/llms-txt|llms.txt]]: a plain-text manifest that lists a site's pages for LLM agents.
- [[glossary/materialized-path|Materialized Path]]: stores the full ancestor path of each tree node as a string, enabling fast subtree queries with simple LIKE or prefix operators.
- [[glossary/materialized-view|Materialized view]]: stores the result of a query as a physical table that can be indexed and queried directly, trading storage and staleness for faster read performance on expensive aggregations.
- [[glossary/mcp|MCP (Model Context Protocol)]]: an open protocol by Anthropic that standardizes how AI agents discover and call external tools, resources, and prompts through a JSON-RPC transport.
- [[glossary/monorepo|Monorepo]]: a single version-controlled repository that contains multiple projects.
- [[glossary/multimodal|Multimodal]]: a model that accepts and processes more than one modality of input (text, images, audio, video, documents) within a single context, enabling cross-modal reasoning.
- [[glossary/optimistic-locking|Optimistic locking]]: detects conflicts at commit time by checking a version counter instead of holding a row lock during the read-modify-write cycle, reducing contention in low-conflict workloads.
- [[glossary/partition-key|Partition Key]]: A partition key is the column or expression used to route each row to a specific database partition or shard, determining data distribution and query routing.
- [[glossary/planner-executor|Planner-Executor]]: Planner-executor splits an LLM agent into a planning phase for task decomposition and an execution phase that carries out each step, reducing errors in both.
- [[glossary/polyrepo|Polyrepo]]: a repository strategy where each project lives in its own version-controlled repo.
- [[glossary/prompt-cache|Prompt Cache]]: lets the model provider reuse KV-cache state from a previous request that shares a common prefix, reducing latency and cost on requests with large repeated context.
- [[glossary/prompt-injection|Prompt injection]]: an attack that smuggles instructions into a model through untrusted input.
- [[glossary/query-plan|Query Plan]]: the execution strategy chosen by the database query planner, describing which indexes, joins, and scan methods will be used to return rows for a SQL statement.
- [[glossary/query-planner|Query Planner]]: The query planner analyzes a SQL statement and picks the lowest-cost execution plan based on table statistics and available indexes.
- [[glossary/rag|RAG (Retrieval-Augmented Generation)]]: RAG augments an LLM's response by retrieving documents from an external store at inference time and injecting them into the prompt as grounding context.
- [[glossary/read-replica|Read replica]]: a continuously-synchronized copy of a primary database that serves SELECT queries, offloading read traffic and enabling horizontal read scaling.
- [[glossary/replication-lag|Replication Lag]]: Replication lag is the delay between a write committing on the primary and that change becoming visible on a replica, measured in WAL bytes or elapsed time.
- [[glossary/reranker|Reranker]]: a cross-encoder model that scores query-document pairs for relevance after first-stage retrieval, improving precision by considering full query-document interaction.
- [[glossary/retrieval-augmented-generation|Retrieval-augmented generation (RAG)]]: a pattern that retrieves relevant context and passes it to a model at generation time.
- [[glossary/role-priming|Role Priming]]: Role priming assigns a persona or professional identity to an LLM at conversation start, shaping its tone, knowledge emphasis, and response style.
- [[glossary/schema-validated|Schema-Validated Output]]: LLM-generated content checked against a JSON Schema or Pydantic model before use, ensuring type safety, required fields, and constrained values.
- [[glossary/secondary-index|Secondary index]]: a data structure that maps non-primary-key column values to row locations, enabling fast lookups and sorted scans without reading the full table.
- [[glossary/semantic-cache|Semantic cache]]: stores LLM responses keyed by the embedding of the prompt and returns a cached result when a new prompt is semantically similar above a threshold.
- [[glossary/semantic-versioning|Semantic versioning (SemVer)]]: a version-number scheme of MAJOR.MINOR.PATCH that signals breaking changes.
- [[glossary/sequential-scan|Sequential Scan]]: reads every row in a table from disk in physical storage order; the fallback scan method when no index is cost-effective for the query.
- [[glossary/server-component|Server component]]: a React Server Component that renders exclusively on the server, ships zero client-side JavaScript for itself, and can access server resources directly.
- [[glossary/sharding|Sharding]]: Sharding partitions a database horizontally across multiple independent nodes, distributing data and query load so that no single node holds the full dataset.
- [[glossary/side-effect|Side effect]]: any state change a function makes outside its return value.
- [[glossary/specificity|Specificity]]: the CSS weight calculation that determines which rule wins when two rules target the same element and the same property.
- [[glossary/ssg|SSG (Static Site Generation)]]: SSG pre-renders all pages to static HTML files at build time, enabling instant delivery from a CDN with no server-side computation per request.
- [[glossary/ssr|SSR (Server-Side Rendering)]]: SSR generates HTML on the server for each request, delivering fully-formed markup to the browser before any JavaScript executes.
- [[glossary/stop-sequence|Stop Sequence]]: a token or string passed to an LLM API that halts generation when the model produces it, allowing callers to bound output length or delimit structured sections.
- [[glossary/structured-output|Structured output]]: model output constrained to a schema, usually JSON.
- [[glossary/structured-prompt|Structured Prompt]]: A structured prompt instructs an LLM to produce output in a machine-readable format such as JSON or XML, enabling reliable parsing without brittle regex.
- [[glossary/suspense-boundary|Suspense boundary]]: a React component that catches asynchronous loading states in its subtree and renders a fallback UI until the data or code chunk is ready.
- [[glossary/system-message|System Message]]: The system message is the privileged prompt sent to an LLM before conversation begins, establishing its role, constraints, output format, and guardrails.
- [[glossary/system-prompt|System prompt]]: the top-level instructions that frame a model's behavior for a session.
- [[glossary/temperature|Temperature]]: a sampling parameter that scales the probability distribution over next-token predictions; lower values produce more deterministic output and higher values increase variability.
- [[glossary/throttling|Throttling]]: Throttling enforces a minimum interval between function calls, limiting execution rate during sustained high-frequency events.
- [[glossary/token|Token]]: the atomic unit of text that a language model processes; one token is roughly 4 characters or 0.75 words of English prose.
- [[glossary/tool-call|Tool Call]]: a structured request emitted by a language model asking the caller to execute a defined function and return its result, enabling models to interact with external systems.
- [[glossary/tool-result|Tool Result]]: A tool result is the structured response injected into the LLM's context after a tool call, providing the observation needed to continue reasoning.
- [[glossary/tool-use|Tool use]]: the capability of a language model to request execution of external functions, APIs, or services during generation and incorporate the results into its response.
- [[glossary/top-p|Top-p (nucleus sampling)]]: restricts token sampling to the smallest set of candidates whose cumulative probability exceeds p, preventing low-probability tail tokens without eliminating variability.
- [[glossary/transaction-isolation|Transaction isolation]]: defines how and when changes made by one database transaction become visible to others, with four standard levels trading consistency for concurrency.
- [[glossary/tree-shaking|Tree shaking]]: a dead-code elimination technique that removes unused exports from an ES module dependency graph before bundling, reducing final bundle size.
- [[glossary/twelve-factor|Twelve-factor app]]: a methodology of twelve rules for building portable, deployable web services.
- [[glossary/vacuum|Vacuum]]: the PostgreSQL maintenance command that reclaims storage from dead tuples created by MVCC, updates the visibility map, and prevents transaction ID wraparound.
- [[glossary/vector-similarity|Vector similarity]]: a numeric measure of how close two embedding vectors are in high-dimensional space, used to rank documents by semantic relevance to a query.
- [[glossary/viewport|Viewport]]: the visible area of a web page in the browser window; understanding its dimensions and how they relate to CSS units is fundamental to responsive design.
- [[glossary/virtual-dom|Virtual DOM]]: an in-memory representation of the real DOM tree that a framework diffs against the previous snapshot to compute the minimal set of real DOM mutations needed.
- [[glossary/wal-mode|WAL Mode]]: Write-Ahead Logging is the durability mechanism where changes are written to a sequential log before data pages are modified, enabling crash recovery, point-in-time restore, and streaming replication.
- [[glossary/web-vitals|Web Vitals]]: Web Vitals are Google's set of real-user experience metrics, with the Core Web Vitals subset (LCP, INP, CLS) used as ranking signals in Google Search.
- [[glossary/write-ahead-log|Write-ahead log]]: a sequential record of every database change written to disk before the change is applied to data files, providing crash recovery and the foundation for replication.
- [[glossary/write-amplification|Write Amplification]]: Write amplification is the ratio of physical bytes written to storage versus logical bytes, driven by indexing, WAL, copy-on-write, and compaction overhead.

## Related MOCs

- [[seo/index|SEO]]
- [[ai-agents/index|AI Agents]]
- [[coding/index|Coding]]
- [[file-organization/index|File Organization]]
- [[meta/index|Meta]]