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:

ADR Title How
ADR-0001 Explicit Opt-In No side effects on import; all components are explicitly created
ADR-0006 Cross-Cutting Decorators Convenience decorators where appropriate
ADR-0008 Free-First Zero external dependencies where practical
ADR-0009 Core Principles Modular, composable components with contracts over implementations
ADR-0017 Agent-Neutral No agent framework dependency
ADR-0020 Capability-Protocol Separation Transport-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