# Organizational Intelligence and AI Agent Glossary ## Organizational Intelligence The ability of an organization to coordinate people, AI agents, workflows, memory, governance, and learning systems into a continuously improving system capable of making decisions and executing work. ## Layers ### Governance Layer The Governance Layer sets the rules, safety boundaries, and approval standards that every agent in the swarm must respect. ### Planning + Coordination Layer The Planning and Coordination Layer turns goals into plans and routes each step to the right agent, person, workflow, or tool. ### Memory + Recursive Learning Layer The Memory and Recursive Learning Layer remembers context, decisions, and lessons, and feeds them back so every loop starts smarter than the last. ### Execution Layer The Execution Layer does the actual work, using specialist expertise, tools, functions, and connected services to turn plans into real outputs. ### Delivery Layer The Delivery Layer packages finished work into clear, usable deliverables and hands them off to the people or systems that need them. ### Simulation Layer The Simulation Layer models possible outcomes and tests strategies before the swarm commits real resources to them. ## Swarm Loop stages ### Define Define is where a mission becomes concrete. The swarm states the outcome it is trying to reach, the constraints it must respect, and the signals that will prove the work succeeded. Nothing is routed or executed until the mission is clear enough to act on. ### Plan Plan converts the mission into a structured set of tasks, dependencies, and fallbacks. It decides what has to happen, in what order, and what to do when a step fails. ### Route Route matches each task to the right resource at the moment it runs. It weighs cost, speed, and quality, then assigns work to the agent, model, or tool most likely to succeed. ### Execute Execute is where plans become real outputs. Specialist and tool-using agents carry out their assigned tasks, calling tools, functions, and external services within the boundaries governance has set. ### Evaluate Evaluate judges whether the work meets the success criteria. It critiques outputs, catches errors, and either improves them through feedback or sends them back for rework. ### Govern Govern enforces the rules the swarm must respect. It can approve, block, or halt work, require human sign-off, and keep the whole system inside its safety, cost, and policy limits. ### Deliver Deliver turns finished work into clear, usable deliverables and hands them off. It formats outputs for their audience and routes them to the right destination. ### Learn Learn writes outcomes, decisions, and lessons back into memory. It is what turns a single run into institutional knowledge the next mission can build on. ## Agent roles ### Governor Agent The Governor Agent enforces system policies, risk limits, and safety rules. It can approve, block, or halt any agent action before it executes, and it decides when a human must sign off. ### Monitor Agent The Monitor Agent watches the whole system for performance, cost, safety, loops, and failures. It surfaces problems early and can trigger a halt when something goes wrong. ### Planner Agent The Planner Agent converts goals into executable plans. It selects strategies, defines workflows, and adapts the plan when conditions change. ### Routing Agent The Routing Agent selects the best agents, models, tools, and workflows for each task, optimizing for cost, speed, and quality. ### Manager Agent The Manager Agent orchestrates execution across specialized agents. It manages dependencies, tracks progress, and ensures objectives are completed. ### Memory Agent The Memory Agent manages short-term, long-term, and compressed memory. It controls what the system stores and what it retrieves when needed. ### Evaluator Agent The Evaluator Agent critiques outputs and improves them through feedback loops. It checks work against the brief and either refines it or sends it back. ### Specialist Agent Specialist Agents are domain-specific workers operating under a manager. Each focuses on a single area of expertise, such as support, coding, research, sales, or legal review. ### Tool-Using Agent The Tool-Using Agent executes explicit functions through tools and APIs. It reads, writes, and acts on external systems on behalf of the swarm. ### Function-Triggering Agent The Function-Triggering Agent emits structured outputs that trigger predefined external functions. It produces the exact payload a downstream system expects. ### Connector Agent The Connector Agent uses sandboxed extensions to access external services safely. It reaches outside systems without exposing the core swarm to them. ### Delivery Agent The Delivery Agent packages outputs into polished deliverables for humans or clients. It formats finished work for its audience and hands it off. ### Simulation Agent The Simulation Agent models possible outcomes and tests strategies before execution. It runs scenarios so the swarm can choose the strongest path first. [Browse the interactive glossary](/glossary) --- Canonical HTML: https://agenticswarm.com/glossary