Skip to main content

COMPEL Glossary / agent-runtime

Agent runtime

The execution substrate that hosts the agent loop, its tool calls, state, and recovery logic — e.g., LangGraph, CrewAI, AutoGen, or OpenAI Agents SDK.

What this means in practice

Choice of runtime is an architectural commitment: it determines how state graphs are expressed, how tool-call authorization is enforced, and how the agent can be observed.

Synonyms

agent execution runtime , agent substrate

See also

  • State graph — An agent execution modeled as a directed graph of states and transitions — each node is a reasoning or tool-call step, each edge is a conditional transition.
  • Agent loop — The iterative pattern in which a model plans, calls tools, observes the results, and replans — continuing until a goal is met, an iteration budget is exhausted, or a human intervenes.
  • Agentic platform — A shared enterprise platform providing agent runtime, tool registry, memory stores, safety layer, observability, and evaluation — so individual agentic use cases inherit these components rather than build them.
  • Tool registry — An authoritative inventory of tools — with schemas, permissions, owners, deprecation state, and audit log — that an agent may call.

Related articles in the Body of Knowledge