simonw / simonw/llm

Dynamic filtering of fragments (aka RAG)

Open
#899 3 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

embeddings fragments
Dominant language
Python
Stars
12.5k
Forks
998
Avg merge
3d 13h
Merged PRs (30d)
10

Description

There are many ways of doing RAG but the gist of it is: based on the user prompt, select which fragments to give to the LLM.

I wanted to create a small RAG and thought of the fragment loader before realizing it's not meant for that.

The "take prompt + (lazy?) fragment collection and return a list of fragments" seems like a generic and powerful abstraction, and it could implement:

  • RAG – in any form people want to implement it
  • Filter down fragments to fit smaller context windows
  • Filtering according to file types (CSS, Python, ...)

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

The issue proposes a prompt-driven fragment collection abstraction for RAG and context filtering, but names no files, tests, or entry points. Start by surveying the existing fragment loader and repository structure, then clarify the API and scope; done requires an agreed design and implementation plan.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
ai
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.