AI Agent Reliability
The measurable degree to which an agent behaves as intended, stays within authority, and recovers safely.
Read the concept →Canonical concepts
Agent Reliability Engineering Concepts
A shared vocabulary for engineering autonomous systems whose behavior must remain measurable, bounded, observable, recoverable, and accountable.
Reliability Surface
The behavioral dimensions and engineering controls that determine whether an agent can be trusted in production.
Read the concept →Reliability Debt
The accumulated risk created when capability, complexity, or authority grows faster than reliability engineering.
Read the concept →Progressive Autonomy
Authority that expands with demonstrated reliability and contracts when reliability deteriorates.
Read the concept →Agent Authority
The explicit, scoped set of actions an agent may take and the data those actions may touch.
Read the concept →Agent Governance
The organizational structures, technical policies, and accountability mechanisms that govern agent behavior.
Read the concept →Agent Reliability Controls
The concrete engineering mechanisms that implement and protect the Reliability Surface.
Read the concept →Reliability Maturity
A five-level model for how systematically agent reliability engineering has been implemented.
Open the maturity model →Missing a term?
New concepts are added through community review, not by editorial fiat.
Propose a concept →How the concepts fit together
The Reliability Surface describes the behavioral properties and engineering controls a production agent needs. Reliability Debt appears when capability, complexity, or authority grows faster than those controls. Progressive Autonomy connects the two by requiring evidence before authority expands and by contracting authority when reliability deteriorates. Together, these concepts give teams a practical vocabulary for designing, evaluating, operating, and governing autonomous AI systems.