CoderGPT is a versatile command-line interface (CLI) designed to enhance coding workflows. It leverages the capabilities of Large Language Models (LLM) and Generative Pre-trained Transformers (GPT) to assist developers in various tasks such as commenting, optimizing, documenting, and testing their code. This tool integrates seamlessly with langchain, providing a powerful backend for code generation and modification.
- OpenAI [
gpt-3.5-turbo,gpt-4,gpt-4-turbo-preview(default)] - Google [
gemini-pro] - Anthropic [
claude-2.1]
Before you begin using CoderGPT, you must set the OPENAI_API_KEY, GOOGLE_API_KEY and ANTHROPIC_API_KEY environment variables on your machine. This key enables authentication with the OpenAI and Google APIs, which are essential for the language model's operation.
export OPENAI_API_KEY='your-api-key-here'
export GOOGLE_API_KEY='your-api-key-here'
export ANTHROPIC_API_KEY='your-api-key-here''Ensure that you replace your-api-key-here with your actual OpenAI API key to enable the full functionality of CoderGPT.
Install CoderGPT easily using pip:
pip install codergptInvoke the CoderGPT CLI with the following syntax:
codergpt [OPTIONS] COMMAND [ARGS]...-v, --verbose INTEGER: Adjust the verbosity level (0 for default, 1 for verbose, 2 for debug).-q, --quiet: Suppress output.--version: Show the version number and exit.--model: Model to use for performing requests.- Available models:
- OpenAI: [
gpt-3.5-turbo,gpt-4,gpt-4-turbo-preview(default)] - Google: [
gemini-pro] - Anthropic[
claude-2.1]
- OpenAI: [
- Available models:
Analyze a package and display its file-language mapping.
codergpt --model <model-name> inspect <path>$ codergpt --model gpt-4 inspect src/codergpt/Provide an explanation for a specific function or class within a package.
codergpt explain <path> [--function <function_name>] [--classname <class_name>]$ codergpt explain src/codergpt/explainer/explainer.py --function explainAutomatically add comments to your code. Choose whether to overwrite the existing file or create a new one.
codergpt comment <path> [--overwrite/--no-overwrite]$ codergpt comment greetings.py --overwriteEnhance your code by optimizing it and adding comments. You can decide to overwrite the original file or save the changes separately.
codergpt optimize <path> [--overwrite/--no-overwrite]$ codergpt optimize example.py --overwriteGenerate test cases for a specified code file, targeting particular functions and/or classes.
codergpt write-tests <filename> [--function <function_name>] [--class <classname>] [--outfile <output_filename>]$ codergpt write-tests example.py --function add --class CalculatorCreate documentation for a given code file by processing and explaining the code.
codergpt document <path> [--outfile <output_filename>]$ codergpt document example.pyThe CoderGPT CLI is developed in Python, utilizing the click library for creating commands. Here's a template for adding a new command:
import click
from codergpt import CoderGPT
coder = CoderGPT()
@click.command()
@click.argument('path', type=click.Path(exists=True))
def new_command(path):
# Implement command logic here
passContributions to CoderGPT are highly encouraged! Please review our contributing guidelines before making pull requests.
CoderGPT is open-sourced under the MIT License. For more details, refer to the LICENSE.md file.
This project was scaffolded using the cookiecutter framework, based on the monarch-project-template. Updates are managed through cruft.
For comprehensive details on the CoderGPT CLI, please refer to the official documentation.