https://github.com/liamconnell/codelabyrinth
Code Labyrinth is a tool for using LLM chatbots in coding workflows. It uses Langchain and PGVector to provide context about the working codebase and relevant libraries.
https://github.com/liamconnell/codelabyrinth
langchain pgvector
Last synced: 7 months ago
JSON representation
Code Labyrinth is a tool for using LLM chatbots in coding workflows. It uses Langchain and PGVector to provide context about the working codebase and relevant libraries.
- Host: GitHub
- URL: https://github.com/liamconnell/codelabyrinth
- Owner: LiamConnell
- Created: 2023-05-22T02:21:02.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2023-05-22T02:31:18.000Z (over 2 years ago)
- Last Synced: 2025-04-14T15:26:45.930Z (10 months ago)
- Topics: langchain, pgvector
- Language: Python
- Homepage:
- Size: 821 KB
- Stars: 8
- Watchers: 1
- Forks: 2
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Code Labyrinth
Code Labyrinth is a tool for using LLM chatbots in coding workflows.
It uses `Langchain` and `PGVector` to provide context about the working codebase and relevant libraries.
## Features
1. Ingest document collections from any `readthedocs` website
2. Ingest all code files from a local directory or github repo and keep them up to date
3. Add relevant documents to the context using vectorstore similarity retrieval of questions asked to a LLM.
4. Choose the appropriate agent for the task - simple Question-Answer agents or Plan-and-Execute agent
5. Operate everything with a convenient user interface
## Setup instructions
1. Start the Postgres PGVector server: `docker compose up`
2. Set up PGVector extension: `python -c "from coder.db import install_extension; install_extension()"`
3. Set up python environment: `virtualenv venv && venv/bin/activate && pip install .`
4. Start backend server: `python -m coder api`
5. Set up frontend environment: `cd ui && npm install`
6. Start frontend (dev) server: `npm run start`