Expand description
Core traits and shared data models for the Rust Deep Agents SDK. This crate keeps the domain primitives lightweight and platform-agnostic so runtimes and integrations can compose them without pulling in heavy deps.
Re-exports§
pub use agent::AgentDescriptor;pub use agent::AgentHandle;pub use agent::PlannerHandle;pub use command::Command;pub use command::StateDiff;pub use events::AgentCompletedEvent;pub use events::AgentEvent;pub use events::AgentStartedEvent;pub use events::EventBroadcaster;pub use events::EventDispatcher;pub use events::EventMetadata;pub use events::PlanningCompleteEvent;pub use events::StateCheckpointedEvent;pub use events::SubAgentCompletedEvent;pub use events::SubAgentStartedEvent;pub use events::TodosUpdatedEvent;pub use events::ToolCompletedEvent;pub use events::ToolFailedEvent;pub use events::ToolStartedEvent;pub use hitl::AgentInterrupt;pub use hitl::HitlAction;pub use hitl::HitlInterrupt;pub use messaging::AgentMessage;pub use messaging::CacheControl;pub use messaging::MessageContent;pub use messaging::MessageMetadata;pub use messaging::MessageRole;pub use messaging::ToolInvocation;pub use persistence::Checkpointer;pub use persistence::CheckpointerConfig;pub use persistence::InMemoryCheckpointer;pub use persistence::ThreadId;pub use tools::Tool;pub use tools::ToolBox;pub use tools::ToolContext;pub use tools::ToolParameterSchema;pub use tools::ToolRegistry;pub use tools::ToolResult;pub use tools::ToolSchema;
Modules§
- agent
- command
- events
- Event system for agent lifecycle tracking and progress broadcasting
- hitl
- Human-in-the-Loop (HITL) types for agent execution interrupts.
- llm
- messaging
- persistence
- Persistence traits for checkpointing agent state between runs.
- prompts
- security
- Security utilities for PII protection and data sanitization
- state
- tools
- Core tool system for AI agents