Skip to content

Repository files navigation

rag

Standalone RAG (Retrieval-Augmented Generation) composition layer for document ingestion, embeddings, and retrieval. The repository is being refactored so reusable capabilities live in dedicated FlossWare repositories such as chunking, storage, and retrieval.

Status

This repository is the RAG composition/application layer. Capability implementations are being extracted into independently reusable repositories.

Install

pip install -e .

Architecture

document
|
v
chunking
|
+----> storage
|
+----> embedding
|
v
retrieval
|
v
evidence / context
|
v
rag composition
|
v
generation

The goal is to keep RAG composition separate from foundational capabilities. Chunking, storage, and retrieval should be independently usable by other applications, agents, and workflows.

FlossWare Engineering Standards

This package complies with the following FlossWare Engineering Standards ADRs:

ADRTitleHow
ADR-0001Explicit Opt-InNo side effects on import; all components are explicitly created
ADR-0006Cross-Cutting DecoratorsConvenience decorators where appropriate
ADR-0008Free-FirstZero external dependencies where practical
ADR-0009Core PrinciplesModular, composable components with contracts over implementations
ADR-0017Agent-NeutralNo agent framework dependency
ADR-0020Capability-Protocol SeparationTransport-independent RAG capabilities

See STANDARDS.md for full compliance details.

License

MIT

About

FlossWare AI Toolkit — standalone, zero-dependency Python package

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages