Core Concepts
An Agent packages:
- a name
- instructions
- a model identifier
- zero or more tools
- zero or more handoffs
- optional output guardrails
A tool is callable code exposed to the model. In aioc, tool execution is never a direct consequence of model output. Tool calls are proposals that still need deterministic authorization.
Handoff
Section titled “Handoff”A handoff lets one agent delegate to another. Internally, aioc models handoffs as runtime-managed tool proposals so they can pass through the same governance boundary as normal tools.
Policy Gate
Section titled “Policy Gate”Policies are deterministic functions that decide whether a proposed action may proceed. Current stable behavior is default deny unless a policy returns allow.
RunRecord
Section titled “RunRecord”RunRecord is the run-level audit artifact emitted by the runtime when recording is enabled. It captures:
- input/output trajectory items
- prompt snapshots
- request fingerprints
- policy decisions
- optional guardrail decisions
- metadata and redacted context snapshot