Guardrails
Rules and constraints that prevent an agent from taking harmful or unauthorized actions.
Why it matters
Agents that take real-world actions need boundaries. Guardrails define the safe operating envelope.
In practice
Our Chat Agent can answer questions about services and pricing but cannot make promises, share internal data, or act outside its scope.
Related terms
Governance
Policies, controls, and processes that ensure responsible agent behavior in production.
Human-in-the-Loop (HITL)
A human reviews and approves specific agent actions before they execute.
Prompt Injection
An attack where external content hijacks an agent into acting against its instructions.
Budget Controls
Spending limits per agent, workflow, or company that prevent runaway API costs.