Module types 

Structs

AgentConfig

Agent configuration controlling the loop.

AgentToolResult

A live result returned from a tool execution.

AllowAllPolicy

Always-allow policy — useful for tests and non-interactive runs.

Enums

AgentEvent

Events emitted by the agent loop, mirroring AgentEvent in TS.

PermissionDecision

Permission outcome for a tool call. Returned by a PermissionPolicy.

Traits

AgentTool

Tool execution trait — analog of AgentTool.execute in TS.

PermissionPolicy

User-supplied permission policy. Implementations may prompt interactively or consult a static allow-list.

Functions

tool_def