langley/codesherpa

A code interpreter ChatGPT plugin.

CodeSherpa

CodeSherpa is a multi-language code-interpreter ChatGPT plugin. It's designed to execute code and run commands locally in a Docker container.

Vector.field.on.a.sphere.animation.mp4

Recent Updates

• May 22, 2023: Refactored README to provide clear and concise instructions for building and running CodeSherpa.
• May 20, 2023: CodeSherpa now supports multiple programming languages, including Python, C++, and Rust.

Quickstart

Prerequisites

Ensure the following software is installed on your system:

• Python 3.10
• Docker

Installation and Running CodeSherpa

Here are the steps to get CodeSherpa up and running swiftly:

Option 1: Using Docker image from Github Packages

# Pull the Docker image
docker pull ghcr.io/iamgreggarcia/codesherpa:latest

# Run the Docker image locally
docker run -p 3333:3333 ghcr.io/iamgreggarcia/codesherpa:latest python3 -c "import localserver.main; localserver.main.start()"

Option 2: Using the repository and Make commands

# Clone the repository
git clone https://github.com/iamgreggarcia/codesherpa.git

# Navigate to the repository directory
cd codesherpa

# Build the Docker image using Make
make build-docker

# Run the Docker image locally
make run-docker-localserver

Option 3: Using the repository and Docker commands

Instead of Make commands, you can use the following Docker commands directly:

# Clone the repository
git clone https://github.com/iamgreggarcia/codesherpa.git

# Navigate to the repository directory
cd codesherpa

# Build the Docker image
docker build -t codesherpa .

# Run the Docker image locally
docker run -p 3333:3333 codesherpa python3 -c "import localserver.main; localserver.main.start()"

Whichever option you choose, CodeSherpa will be accessible at localhost:3333.

Connecting CodeSherpa to ChatGPT

1. Navigate to the ChatGPT UI, and access the plugin store.
2. Select "Develop your own plugin".
3. In the plugin URL input, enter localhost:3333. Your ChatGPT should now be able to use CodeSherpa's features.

For more detailed information on API endpoints, usage, contributing, future work, and license, refer to the respective sections below.

API Endpoints

POST /repl

Executes the provided code is the specified language, maintaining the state between requests, and returns the output or error message.

Request Body

Parameter	Type	Description
code	string	The code to be executed.
language	string	The programming language of the code to be executed.

Supported Languages

Language	Language Parameter
Python	python
C++	c++
Rust	rust

Response

HTTP 200

Parameter	Type	Description
result	string	The output of the executed code.

HTTP 400

Parameter	Type	Description
error	string	A brief description of the error.
returnCode	integer	The return code of the failed process.
command	string	The command that was attempted.
output	string	Any output that resulted from the attempted command.

POST /command

Runs the provided terminal command and returns the output or error message.

Request Body

Parameter	Type	Description
command	string	The terminal command to execute.

Response

Parameter	Type	Description
result	string	The output of the executed command.
error	string	Error message if there is an error.

Usage

To interact with the CodeSherpa API, use HTTP requests to the corresponding endpoints. To execute some code code, send a POST request to the /repl endpoint with a JSON object containing the code key and the language key.

For example, here's a request to plot a vector field on a sphere using Python:

{
  "code": "import numpy as np\nimport matplotlib.pyplot as plt\nfrom mpl_toolkits.mplot3d import Axes3D\n\n# Define the number of vectors along each dimension\nn_vectors = 20\n\n# Define the sphere\nphi = np.linspace(0, np.pi, n_vectors)\ntheta = np.linspace(0, 2 * np.pi, n_vectors)\nphi, theta = np.meshgrid(phi, theta)\nx = np.sin(phi) * np.cos(theta)\ny = np.sin(phi) * np.sin(theta)\nz = np.cos(phi)\n\n# Define the vector field\nvx = np.sin(phi) * np.cos(theta)\nvy = np.sin(phi) * np.sin(theta)\nvz = np.cos(phi)\n\n# Create the figure\nfig = plt.figure(figsize=(8, 8))\nax = fig.add_subplot(111, projection='3d')\n\n# Plot the sphere surface\nax.plot_surface(x, y, z, color='b', alpha=0.3)\n\n# Plot the vector field\nax.quiver(x, y, z, vx, vy, vz, length=0.1, normalize=True)\n\n# Set the aspect ratio\nax.set_box_aspect([1, 1, 1])\n\nplt.savefig('static/images/vector_field.png')",
  "language": "python"
}

And here's ChatGPT send the request to CodeSherpa:

Vector.fields.and.manifolds.-.plot.mp4

To run terminal commands, send a POST request to the /command endpoint with a JSON object containing the command key and the terminal command to execute.

For example:

{
  "command": "echo \"Hello, terminal!\""
}

Contributing

I welcome contributions! If you have an idea for a feature, or want to report a bug, please open an issue, or submit a pull request.

Steps to contribute:

1. Fork this repository.
2. Create a feature branch git checkout -b feature/YourAmazingIdea.
3. Commit your changes git commit -m 'Add YourAmazingIdea'.
4. Push to the branch git push origin feature/YourAmazingIdea.
5. Submit a Pull Request.

Future Work

So I've built CodeSherpa, this Python interpreter-ChatGPT-plugin creature that runs inside Docker like a hamster on a wheel. It's neat, but very much a local(host) affair. Think charming neighborhood coffee shop, not Starbucks.

Scaling this into a full-blown, resource-intensive production version? A tall order. But I'm working on a lightweight version that can ideally be installed by anyone with ChatGPT Plugin access.

It'll swap Docker for an existing compiler API (which can support multiple languages), acting more like a nimble, stateless code sidekick than a full-blown REPL. It'll be a bit more limited, but a start. Again I'm very open to contributions, so if you have an idea, please open an issue or submit a pull request.

License

This project is licensed under the terms of the MIT license.