AIDLC: AI Development Life Cycle. An iterative collaborative process where AI agents and humans work together. It emphasizes elaborating on initial placeholders/ideas, conducting research, discussing requirements, and clearly defining Acceptance Criteria (AC) before formalizing or implementing a task.
Backlog.md: A Markdown-native Task Manager and Kanban visualizer used via npx backlog. It is the source of truth for all project and task management tracking in this project.
AC: Acceptance Criteria. Clear, testable conditions that a task must meet to be considered Done.
MCP: Model Context Protocol. The protocol used by the agent harness to interface with tools (e.g., the backlog MCP).