Clear explanations of agent identity, delegation, runtime authorization, tool security and evidence for consequential actions.
Useful security language begins with the action, the authority behind it and the boundary that can stop it.
Selected paths
A foundational guide to identity, delegated authority, runtime control and evidence.
Explore this path Resources / Technical GuideA practical guide to tool trust, delegated authority, action-level control and execution evidence.
Explore this path Resources / Technical GuideWhy a decision at task start cannot govern every later tool call or API action.
Explore this pathAll Resources
A working perspective
Start with “What is autonomous authority?” to distinguish an authenticated actor from a permitted outcome. The agent identity guide explains attribution. Delegation and runtime-authorization resources then show how a grant follows an agent chain and reaches the action boundary.
These educational resources explain ProofGrid capabilities through security concepts and public architecture examples while keeping private interfaces and implementation details protected.
Architecture reviews fail when identity, access, authority, policy and verification are used interchangeably. The glossary gives engineers, operators and decision makers a concise shared vocabulary for defining protected actions and reviewing decisions.
Resources / Discussion