Module state

Expand description

Direct re-exports from echo_state::memory.

Modules§

checkpointer: Short-term thread state persistence — concrete implementations
conversation: Conversation persistence — concrete implementations and helper functions
embedder: HTTP embedding client (OpenAI-compatible)
embedding_store: 向量增强 Store（EmbeddingStore）
snapshot: Agent 状态快照与回滚
sqlite_conversationsqlite: SQLite conversation persistence implementation
sqlite_storesqlite: SQLite persistent Store (FTS5 full-text search + optional vector search)
store: Long-term memory Store — concrete implementations

Checkpoint: Snapshot of a single conversation state
Conversation: Complete conversation record
ConversationFilter: List filter criteria
ConversationMeta: Conversation list entry (lightweight, no messages or summary)
EmbeddingStore: 向量增强 Store：在任意 Store 外包装余弦相似度语义检索
FileCheckpointer: JSON file-based persistent Checkpointer
FileStore: JSON file-based persistent Store
HttpEmbedder: OpenAI-compatible HTTP embedding client
InMemoryCheckpointer: In-process memory Checkpointer, state is lost on restart, suitable for tests
InMemoryStore: In-process memory Store, no persistence, suitable for testing and short-lived use
NewConversation: Parameters for creating a new conversation
SearchQuery: Unified search request
SnapshotManager: 管理状态快照的环形缓冲区
SqliteConversationStore: SQLite conversation persistence Store
SqliteStore: SQLite persistent Store with FTS5 full-text search and optional vector search
StateSnapshot: 单次状态快照，记录某一时刻的完整对话历史
StoreItem: A single record in the Store
StoredMessage: Persisted message (independent from LLM Message, with persistence fields)
ThreadState: Thread-level runtime state.

project_message: Project a single runtime Message to a transcript record.
project_messages: Project runtime Message list to persistable transcript records.