Crate terraphim_agent_evolution 

Terraphim Agent Evolution System

A comprehensive agent memory, task, and learning evolution system that tracks the complete development and learning journey of AI agents over time.

Core Features

• Versioned Memory: Time-based snapshots of agent memory states
• Task List Evolution: Complete lifecycle tracking of agent tasks
• Lessons Learned: Comprehensive learning and knowledge retention system
• Goal Alignment: Continuous tracking of agent alignment with objectives
• Evolution Visualization: Tools to view agent development over time

Architecture

The evolution system consists of three core tracking components that work together:

• Memory Evolution: Tracks what the agent remembers and knows
• Task List Evolution: Tracks what the agent needs to do and has done
• Lessons Evolution: Tracks what the agent has learned and how it applies knowledge

All components use terraphim_persistence for storage with time-based versioning.

Re-exports

pub use error::*;

pub use evolution::*;

pub use integration::*;

pub use lessons::*;

pub use llm_adapter::*;

pub use memory::*;

pub use tasks::*;

pub use viewer::*;

Modules

error

Error types for agent evolution system

evolution

Core agent evolution system that coordinates all three tracking components

integration

Integration module for connecting workflows with the agent evolution system

lessons

Agent lessons learned evolution with comprehensive learning management

llm_adapter

LLM adapter for agent evolution system

memory

Agent memory evolution tracking with time-based versioning

tasks

Agent task list evolution with complete lifecycle tracking

viewer

Agent evolution viewer for visualizing agent development over time

workflows

AI Agent workflow patterns implementation

Type Aliases

AgentId

Agent identifier type

EvolutionResult

Result type for agent evolution operations

LessonId

Lesson identifier type

MemoryId

Memory item identifier type

TaskId

Task identifier type