A LangChain-inspired framework for building LLM applications in Rust. Supports OpenAI, Agents, Tools, Memory, Chains, RAG, BM25, Hybrid Retrieval, LangGraph,…
Read layer for the rag-rat workspace: graph/impact/symbol/tree queries, repo-memory reads and evidence resolution, pagerank, and orientation primitives.
The work ledger for ikigai: NAMED ledgers — one named graph and one capability each — holding items, comments, labels, links and claims as RDF in a durable…
Deterministic, field-weighted lexical retrieval over the version-keyed tool catalog: the ranking index behind shortlist advertisement and its CI-gating…
Explainable retrieval for MemoryWhale: a MemoryEngine interface, a built-in scorer that ranks memories with per-signal reasons, and an optional MemPalace…
One discovery/auth convention for Right Suite local services: service.json, ranking and takeover, current-user-only Unix socket / named pipe, per-app request…
Deep memory for agents over large datasets: index JSONL, CSV, SQLite or any database export once, answer with a few ranked, cited records in about five hundred…
Prospect, the full-text search of the Smeltery framework: searchable models, the database driver (SQLite FTS5, PostgreSQL tsvector, MySQL FULLTEXT), ranking,…