Welcome to the official repository for Lux assistant tools. Here, you will find a collection of tools designed to integrate seamlessly with the Lux assistant.
You just need to download the tool ZIP File you want, go on the configuration page on your Lux-Interface and import the ZIP files you want to use.
Follow these steps:
-
Create
tools_list.py
:This is where you will put your function(s) of your tool as well as the description(s) which will enable you to trigger the tool via the similarity between the user prompt and your description(s).
from kernel.tools.tools_functions.tools_action.your_folder.your_script import function_1, function_2 tools = { "function_1": { "description": "description pour activer une fonction de votre outil" if LANGUAGE == 'fr' else "description to activate a function of your tool", "function": your_function_name }, "function_2": { "description": "autre description pour activer une fonction de votre outil" if LANGUAGE == 'fr' else "other description to activate a function of your tool", "function": your_other_function_name }, }
-
(if needed) Create
requirements.txt
:
List any external libraries required by your tool that are not already included in Lux.For example :
nltk==3.9.1
-
Create
menu.py
:Define a dictionary containing the global description of your tool. This will be used to activate and access your tool.
{ "name": "name of your tool", "description_fr": "la description en français de l'outil", "description_en": "the English description of the tool", "emoji": "optional emoji" }
-
Create a Folder:
- For action tools, place your scripts in the
tools_action
folder. - For response tools, place your scripts in the
tools_response
folder.
- For action tools, place your scripts in the
-
(if needed) Create
select_tool.py
:If your tool needs to take user input as a parameter, create a
select_tool.py
file listing the names of the functions that require user input.["function_name"]
---Example for a tool that takes user input---
On the
select_tool.py
you put: ["vocal_note"]On the
tools_list.py
you put:from kernel.tools.tools_functions.tools_action.take_note import vocal_note tools = { "vocal_note": { "description": "prends note" if LANGUAGE == 'fr' else "take note", "function": vocal_note }, }