CLI tool for the Luca assistant
Project description
LUCA CLI Client
Overview
The LUCA CLI Client is a command-line interface for your LUCA assistant.
It allows you to interact with an assistant that lives in your terminal. This is a non-intrusive user experience, that means you still have complete control over your terminal. But the assistant is always a command away.
Capabilities
We have designed the system to be able to:
- Retrieve and search relevant research papers from ArXiv.
- Retrieve experiments logged in a Weights & Biases project.
- Generate and execute Python and bash commands.
With these capabilities, you can use the assistant to:
- Generate reports that theorize and summarize your research experiments.
- Generate a project plan to tackle a new research problem.
- Brainstorm, ideate generate new hypotheses based on your current experiments.
Pre-installation
Before installing the LUCA CLI, please make sure you have created an account on the LUCA platform.
Additionally, make sure you have exported the SERVER_URL environment variable to your local machine.
export SERVER_URL="http://<your-server-ip>:8000"
You can find the SERVER_URL in your LUCA account dashboard.
Installation
pip install luca-cli
Usage
luca --help
As soon as you pip install the package, please run the following command under any $ROOT directory:
luca init
This will initialize the assistant and create a knowledge base $ROOT/.luca/kb.txt.
This knowledge base will be updated with new information as you use the assistant.
After the initialization, you can start interacting with the assistant by just typing your prompt:
luca "Research papers on reinforcement learning."
If you are using Weights & Biases to log your experiments, you can set your
W&B API key and entity name as environment variables and re-initialize the assistant:
export WANDB_API_KEY="your-wandb-api-key"
export WANDB_ENTITY="your-wandb-entity"
luca init
This will update the assistant and allow the assistant to access your W&B experiments. You can then do cool things like:
luca "Export a powerpoint report of all the experiments in my wandb project <your-wandb-project-name>?"
You can create reports that theorize and summarize your experiments and help you
analyze the experiments to come up with better hypotheses for the next set of experiments.
Any file the assistant creates will be synced to your local machine and saved
in under the $ROOT/.luca/artifacts directory.
We plan to significantly expand the set of capabilities of the assistant with each new release.
Please provide your unfiltered thoughts, suggestions and feedback to us by using the luca feedback command.
luca feedback "I love the assistant!"
Cheers,
The LUCA team
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file luca_cli-0.1.4.tar.gz.
File metadata
- Download URL: luca_cli-0.1.4.tar.gz
- Upload date:
- Size: 4.5 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.1.0 CPython/3.13.3
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
27a984fd9302f80a85263b7d1996d118c6ae4f7f2fa1e11a3ea0b00e7a6bedca
|
|
| MD5 |
f2833539b4d9fe3e66747a9f1d574c63
|
|
| BLAKE2b-256 |
f1dff0907e660e3d4660b6fe0526aadf6f6a5c08f20a62eedebe97b82788fbaa
|
File details
Details for the file luca_cli-0.1.4-py3-none-any.whl.
File metadata
- Download URL: luca_cli-0.1.4-py3-none-any.whl
- Upload date:
- Size: 4.5 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.1.0 CPython/3.13.3
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
cde40e7b94a23576a4be64036227c7cca9d480b4dd0662ae8be3c2eaf87a7f2a
|
|
| MD5 |
a2d428a841d3d2127286a97b66a9c576
|
|
| BLAKE2b-256 |
97ec6a3cbcfacd380a91a79a5d69d019ed3d8ee04e4f786ce4b26ec043186c85
|