Glossary
This glossary defines key terms used throughout the AI SDLC Playbook to ensure consistent terminology and understanding.
A
AI IDE - An Integrated Development Environment enhanced with artificial intelligence capabilities, such as Cursor or GitHub Copilot-enabled editors. These tools provide AI-powered code generation, completion, and assistance.
AI Tools - Software applications that leverage artificial intelligence to assist in development tasks, including code generation, testing, documentation, and analysis.
F
Functional Requirements - Specifications that define what a system should do from a user's perspective, focusing on expected behaviors, user interface elements, and testable features. Always used in plural form throughout this playbook.
I
IDE Rules - Configuration files that guide AI behavior in development environments. These rules provide system-level guidance to AI assistants and define coding standards, patterns, and conventions.
M
MCP (Model Context Protocol) - A protocol that enables AI models to connect with external data sources and tools, extending their capabilities beyond their training data.
MCP Server - A backend service that implements the Model Context Protocol, providing AI models with access to specific tools, APIs, or data sources.
Memory Bank - A structured documentation system that helps AI assistants maintain project context across sessions through organized Markdown files.
P
Prompt Engineering - The practice of crafting effective prompts to guide AI models toward desired outputs, including techniques like meta-prompting and iterative refinement.
Project Rules - AI IDE rules that are specific to a project, stored in version control and scoped to the codebase.
T
Technical Specification - A detailed document that defines how functional requirements will be implemented technically, including data models, API endpoints, backend architecture, and implementation details.
U
User Rules - AI IDE rules that are global to a user's environment and applied across all projects, typically used for personal preferences and coding style.