Crate remembrall_core 

Modules

config

embed

Embedding generation for memory content.

error

graph

indexer

Incremental code indexer.

ingest

Ingestion pipeline: bulk-import memories from GitHub PRs and markdown docs.

memory

parser

Source code parsers using tree-sitter.

search

Hybrid search: combines semantic (pgvector) + full-text (tsvector) with RRF fusion. TODO: Implement in Spike 1.

Functions

tokio_block_on_embed

Synchronously embed a single text using a blocking thread. Used by test harnesses that need to generate query embeddings from async contexts.