Crate smooth_operator_adapter_postgres 

Postgres + pgvector StorageAdapter — the dogfood backend.

This is the production Postgres implementation of the one storage seam (see docs/STORAGE.md). It mirrors the smooai monorepo’s schema so dogfooding is a swap, not a rewrite:

• OLTP (conversations / participants / messages / sessions): async CRUD over a deadpool_postgres pool, semantics matching the in-memory baseline (conversation idempotency, external-id participant resolve, cursor message paging, session status/counts).
• Checkpoints: smooth-operator’s PostgresCheckpointStore (a synchronous r2d2-pooled store) constructed against the same database — so the engine’s with_checkpoint_store plugs straight in and agent state lives next to the conversations it belongs to.
• Knowledge: a pgvector-backed PgKnowledgeBase (dense HNSW cosine ∪ sparse tsvector BM25 → Reciprocal Rank Fusion). Text→vector goes through the Embedder seam — DeterministicEmbedder by default (reproducible, no network), GatewayEmbedder when a live gateway is configured.
• Memory: a pgvector-backed PgMemory (parity gap Phase 3 / SMOODEV-1470) — persistent, semantic, cross-thread agent memory namespaced by (organization_id, user_id) like the TS ['memories', orgId, userId] store. Implements the core Memory trait; recall is pgvector cosine top-K under an HNSW index, scoped to the namespace. Shares the adapter’s Embedder.

Sharing one database between the async pool and the sync checkpoint store

The OLTP/knowledge slices need async (tokio-postgres + deadpool); the checkpoint slice is a sync trait backed by an r2d2 pool inside smooth-operator. Both are pointed at the same conn_str: we build the async deadpool from it for our own queries, and hand the same string to PostgresCheckpointStore::connect, which stands up its own small r2d2 pool and checkpoints table in that database. Two pools, two driver stacks, one Postgres — the tables coexist (ours from [schema], theirs from their own CREATE TABLE IF NOT EXISTS).

Structs

DeterministicEmbedder

Deterministic, network-free pseudo-embedder.

GatewayEmbedder

OpenAI-compatible /v1/embeddings embedder (the SmooAI LiteLLM gateway).

GatewayReranker

Cross-encoder reranker over the SmooAI gateway’s /v1/rerank endpoint (feature gap G8).

HttpRerankBackend

The real backend: a Cohere/Voyage-style /v1/rerank call over the gateway.

PgConnectorConfigStore

Postgres-backed ConnectorConfigStore over connector_configs.

PgIndexingStore

Postgres-backed IndexingStore over indexing_runs.

PgKnowledgeBase

pgvector knowledge base. Cheap to clone (all fields are Arc/pool handles).

PgMemory

pgvector-backed agent memory, bound to one (organization_id, user_id) namespace. Cheap to clone (all fields are Arc/pool handles + small strings).

PgSettingsStore

Postgres-backed SettingsStore over agent_settings.

PostgresAdapter

Postgres + pgvector storage adapter.

RerankScore

One scored candidate returned by a RerankBackend: the candidate’s index in the request documents array, and its relevance score against the query.

Enums

InputType

Whether an embedding is for a document being stored or a search query.

Constants

DEFAULT_EMBEDDING_DIM

Default embedding dimension (Voyage voyage-3-large shape; mirrors smooai’s knowledge_vectors embedding vector(1024)).

DEFAULT_RERANK_MODEL

Default rerank model requested over the gateway (Cohere-compatible). Distinct from the embedding model and the chat model.

OPENAI_SMALL_EMBEDDING_DIM

Dimension returned by OpenAI text-embedding-3-small.

Traits

Embedder

Turn text into dense vectors. Implementations must return one vector per input string, each of length Embedder::dim.

RerankBackend

A pluggable rerank backend.