An LLM app for understanding your codebase
Configure your .env files based on the .env-example. Yes, I said files. One goes in the root directory and another goes in the frontend directory. This is a hacky workaround, but it's necessary for the time being.
Install the milvus standalone image via dockerhub, then navigate to the docker directory. From there,
run docker compose up -d to spin up your milvus database based on the provided docker-compose.yml file.
Install the python project with poetry install, and use the poetry run mentor load command to load a file, folder,
or
repository into milvus. It can sometimes take a minute or two for the embeddings to become available after being loaded
in.
run poetry run mentor server to start the app. Alternatively, use poetry run mentor prompt to send prompts via the
command line.