MARSHAL LEXICON
Lexicon
Practical definitions for the systems, concepts, and language behind AI agents and GEO.
A
Agent Evaluation
Agent evaluation is the process of testing and measuring whether agents perform tasks correctly, safely, and consistently.
GovernanceAI Agent SystemsAgent Frameworks
Agent frameworks are developer tools and libraries used to build AI agents, multi-agent systems, tool use, memory, and orchestration logic.
ArchitectureAI Agent SystemsAgent Observability
Agent observability is the ability to monitor, inspect, and understand how agents behave in production.
GovernanceAI Agent SystemsAgent Permissions
Agent permissions define what data, tools, actions, and systems an AI agent is allowed to access or control.
GovernanceAI Agent SystemsAgent Runtime
An agent runtime is the execution environment where an AI agent runs, maintains state, calls tools, and completes tasks.
ArchitectureAI Agent SystemsAgentic AI
Agentic AI refers to AI systems that can pursue goals, plan steps, use tools, and act with some level of autonomy.
Core ConceptsAI Agent SystemsAgentic Automation
Agentic automation is automation powered by AI agents that can reason, adapt, and act across multi-step workflows.
Core ConceptsAI Agent SystemsAgentic Workflows
Agentic workflows are multi-step business processes where AI agents plan, decide, act, and adapt as work progresses.
Core ConceptsAI Agent SystemsAI Agent Architecture
AI agent architecture is the design of the components that allow agents to reason, retrieve context, use tools, follow rules, and execute work.
ArchitectureAI Agent SystemsAI Agent Builder
An AI agent builder is a tool or platform that helps users create agents through configuration, prompts, workflows, and integrations.
ArchitectureAI Agent SystemsAI Agent Platforms
AI agent platforms are software environments for creating, deploying, managing, and monitoring AI agents.
Core ConceptsAI Agent SystemsAI Agents
AI agents are software systems that use AI models to understand goals, make decisions, and take actions through tools or connected systems.
Core ConceptsAI Agent SystemsAI Governance
AI governance is the system of policies, controls, accountability, and oversight used to manage AI safely and effectively.
GovernanceAI Agent SystemsAI Guardrails
AI guardrails are rules, constraints, checks, and controls that limit what an AI system can say or do.
GovernanceAI Agent SystemsAI Memory
AI memory is the ability of an AI system to retain useful context across interactions, tasks, users, or workflows.
InfrastructureAI Agent SystemsAI Operations
AI operations refers to the practices used to deploy, monitor, maintain, and improve AI systems in real business environments.
InfrastructureAI Agent SystemsAI Orchestration
AI orchestration is the coordination layer that manages models, agents, tools, data, workflows, permissions, and outputs.
Core ConceptsAI Agent SystemsAI Reliability
AI reliability is the degree to which an AI system performs consistently, accurately, and safely under real operating conditions.
GovernanceAI Agent SystemsAI Tool Use
AI tool use is the capability of an AI system to select and use external tools to complete a task.
ArchitectureAI Agent SystemsAI Workflow Automation
AI workflow automation uses AI to automate repeatable business workflows that require interpretation, judgment, or content generation.
Core ConceptsAI Agent SystemsAPI Integrations
API integrations connect software systems so they can exchange data and trigger actions programmatically.
AutomationAI Agent SystemsAudit Trails for AI Agents
Audit trails for AI agents are logs that record what an agent did, what tools it used, what data it accessed, and what outputs it produced.
GovernanceAI Agent SystemsAutonomous AI Agents
Autonomous AI agents are agents that can complete tasks with limited human intervention after receiving a goal or instruction.
Core ConceptsAI Agent SystemsB
C
CRM Agents
CRM agents are AI agents that interact with customer relationship management systems to support sales, service, and account workflows.
Business Use CasesAI Agent SystemsCustom AI Agents
Custom AI agents are agents designed around a specific company's workflows, data, tools, voice, and business rules.
ArchitectureAI Agent SystemsCustomer Support Agents
Customer support agents are AI agents that help answer customer questions, resolve issues, route tickets, and support service teams.
Business Use CasesAI Agent SystemsD
E
F
H
I
K
L
LLM Agents
LLM agents are AI agents built on top of large language models that can reason through language, use tools, and complete tasks.
Core ConceptsAI Agent SystemsLLMOps
LLMOps is the operational discipline for deploying, monitoring, evaluating, and improving large-language-model applications.
InfrastructureAI Agent SystemsLong-Term Memory for Agents
Long-term memory for agents refers to persistent memory that allows agents to retain information across sessions and over time.
InfrastructureAI Agent SystemsM
MLOps
MLOps is the practice of managing machine learning models through development, deployment, monitoring, and maintenance.
InfrastructureAI Agent SystemsModel Context Protocol
Model Context Protocol, often called MCP, is an emerging standard for connecting AI systems to tools, data sources, and external context.
ArchitectureAI Agent SystemsMulti-Agent Systems
Multi-agent systems are architectures where multiple specialized agents work together to complete a larger task or workflow.
Core ConceptsAI Agent SystemsO
P
R
Reasoning Models
Reasoning models are AI models designed to perform better on complex tasks that require step-by-step analysis, planning, or decision-making.
ArchitectureAI Agent SystemsRetrieval-Augmented Generation
Retrieval-augmented generation, or RAG, is a method where an AI system retrieves relevant information before generating an answer or taking action.
InfrastructureAI Agent SystemsRobotic Process Automation
Robotic process automation, or RPA, uses software bots to perform repetitive digital tasks across applications.
AutomationAI Agent SystemsS
Sales Agents
Sales agents are AI agents that support sales workflows such as prospect research, lead qualification, follow-up, CRM updates, and pipeline management.
Business Use CasesAI Agent SystemsSemantic Search
Semantic search retrieves information based on meaning and intent rather than exact keyword matching.
InfrastructureAI Agent SystemsT
Task Decomposition
Task decomposition is the process of breaking a larger task or goal into smaller executable steps.
ArchitectureAI Agent SystemsTool-Using AI
Tool-using AI refers to AI systems that can interact with external tools such as APIs, databases, browsers, CRMs, calendars, and business applications.
ArchitectureAI Agent Systems