Skip to main content

VibeSpec Authoritative Glossary

Core Methodology Terms​

Acceptance Criteria​

Specific, measurable conditions that define when a feature or requirement is complete and acceptable for delivery.

Agent​

A specialized AI role with defined responsibilities, explicit boundaries, and specific workflows within the VibeSpec system.

Agent Coordination​

The systematic process of managing multiple specialized agents through explicit hand-off protocols and clear authority boundaries.

Append-Only Pattern​

A data storage approach where historical entries are never deleted or modified, only new entries are added, preserving complete audit trail.

Approval Gate​

A required checkpoint where user must explicitly approve specifications, changes, or implementations before proceeding to the next phase.

Architect Agent​

The VibeSpec agent responsible for planning, design, specification creation, and architectural decision-making with authority over all design decisions.

Architecture Specification​

A formal document defining the technical design, system structure, technology choices, and implementation constraints for a project.

Authority Domain​

The specific area of responsibility where an agent has final decision-making power and cannot be overridden by other agents.

Coder Agent​

The VibeSpec agent responsible for implementation, coding, and pattern application with authority over all implementation decisions within approved specifications.

Context Preservation​

Maintaining relevant project information, decisions, and state across development sessions through persistent file storage.

Debugger Agent​

The VibeSpec agent responsible for problem analysis, root cause identification, and fix implementation with authority over all debugging and resolution decisions.

Explicit Everything Principle​

Core VibeSpec principle requiring all state, decisions, context, and behavior to be stored in human-readable files rather than hidden or implicit systems.

Hand-off Protocol​

Structured process for transferring control and context between agents to ensure continuity and prevent information loss.

Memory​

Persistent, file-based storage of project context, decisions, patterns, and learning that survives across development sessions and enables continuous improvement.

Memory First Principle​

Core VibeSpec principle requiring always reading existing memory before making decisions and always updating memory after significant decisions or discoveries.

Pattern​

A reusable solution to a common problem, documented in memory for future application and sharing across projects.

Product Specification​

A formal document defining what to build, including user requirements, business value, and acceptance criteria for a project or feature.

Prompt-as-Interface​

A design pattern where natural language prompts serve as the primary interface for interacting with AI agents, replacing traditional graphical user interfaces.

Refusal Protocol​

Systematic approach for agents to decline unsafe, inappropriate, or impossible requests while suggesting safe alternatives.

Reviewer Agent​

The VibeSpec agent responsible for quality assurance, safety compliance, and specification verification with authority over all quality and compliance decisions.

Safety Governance​

A framework of absolute, non-negotiable rules that override all other instructions to ensure secure, reliable, and ethical development practices.

Safety Override Principle​

Core VibeSpec principle stating that safety rules override all other instructions, including direct user requests, and security cannot be compromised for any reason.

Spec-Driven Development​

A development methodology where all features and changes require approved specifications before implementation, ensuring thoughtful design and clear traceability.

Test Agent​

The VibeSpec agent responsible for validation, testing, and coverage analysis with authority over all testing strategy and implementation decisions.

Traceability​

The ability to trace any code, feature, or decision back to its originating specification, requirement, or architectural decision.

Vibe Coding​

A development methodology where human developers collaborate with AI agents through explicit protocols, persistent memory, and structured workflows to create software.

VibeSpec​

An agentic, memory-driven, spec-first development operating system that enables vibe coding through file-based protocols, multi-agent coordination, and persistent memory.

File System Terms​

Memory Files​

Persistent storage files in memory/ directory that capture project context, decisions, and learning including decisions.md, mistakes.md, patterns.md, and project.json.

Spec Files​

Formal specifications stored in specs/ directory that define what to build and how to build it, including product.md, architecture.md, and acceptance.md.

System Files​

Core VibeSpec configuration files in .vibespec/ directory that define system behavior, workflows, and safety rules including system.md, workflow.md, and safety.md.

Workflow Terms​

Initialization Sequence​

Required startup process where VibeSpec loads all system files, agents, memory, and specs to establish operational context before beginning work.

Task State Transition​

The progression of tasks through defined states from backlog to active to done as they move through the development workflow.

Quality Terms​

Agent Specialization​

The division of development responsibilities among specialized agents that focus deeply on specific domains rather than attempting to handle all tasks generically.

Authority Hierarchy​

The structured system of decision-making power where safety governance has absolute authority, followed by specification authority, implementation authority, quality authority, validation authority, and resolution authority.

Boundary Respect​

The practice of agents staying within their defined authority and expertise areas while deferring appropriately to other agents for decisions outside their domain.

Consistency Requirements​

The mandate that all VibeSpec terminology, processes, and quality standards remain uniform across all documentation, implementations, and team usage.

Context Transfer​

The complete communication of relevant information, decisions, and state when transitioning work between agents or development sessions.

Decision Documentation​

The systematic recording of all architectural, technical, and process decisions with rationale, alternatives considered, and implementation implications in memory files.

Memory Integration​

The practice of all agents accessing, referencing, and updating project memory to maintain continuity and enable continuous improvement across development sessions.

Quality Gate​

A validation checkpoint where specific quality criteria must be met before work can proceed to the next phase of development.

Systematic Coordination​

The organized management of multiple agents through formal protocols, clear authority boundaries, and structured communication to ensure comprehensive development coverage.

Implementation Terms​

Agent Activation​

The process of establishing a specific agent context, responsibilities, and operational framework that guides AI behavior according to that agent's expertise and authority.

File-Based Protocol​

The VibeSpec approach of using human-readable files to store all system state, configuration, and communication rather than relying on hidden or proprietary formats.

IDE-Agnostic​

The characteristic of VibeSpec working with any development environment capable of file operations, without requiring specific tools or integrations.

Multi-Agent Architecture​

The VibeSpec system design that divides development responsibilities among five specialized agents (Architect, Coder, Reviewer, Test, Debugger) with clear coordination protocols.

Persistent Context​

The maintenance of project information, decisions, and learning across multiple development sessions through file-based storage that survives system restarts and time gaps.

Sequential Workflow​

The most common VibeSpec development pattern where agents work in order (Architect → Coder → Reviewer → Test → Debugger) with formal hand-offs between phases.

Specification Authority​

The hierarchical system where the Architect Agent has final authority on requirements and design decisions that cannot be overridden by implementation preferences.

Structured Prompts​

Formatted natural language instructions that follow consistent patterns to ensure predictable and effective communication with AI agents.

Glossary Authority: This glossary serves as the single source of truth for all VibeSpec terminology and must be used consistently across all documentation, implementations, and communications.

Usage Requirements: All authors, contributors, and users must reference these exact definitions to ensure consistency and prevent terminology confusion.

Update Protocol: New definitions require Architect Agent approval, changes to existing definitions require user approval, and all updates must be documented in memory/decisions.md.