Codebundle generator

Codebundle Generator is the easiest way to generate your Robot Framework based CLI Codebundles. Get some examples here. with a user-friendly web interface.

Supercharged by ChatGPT.

Disclaimer

This is only a simple prototype, the code is untested and unstable. Workspace state is ephemeral. Inputs are not validated.

Table of Contents

• Features
• Functionalities
• Coming soon
• Installation guide
• Features suggestions
• License

Features

Codebundles generation made even easier

Writing Codebundles using Robot framework is very easy and intuitive. But why not to use GUI to save some time familiarizing with the syntax and exploring Keywords documentation. Don't burn your time on writing everything from scrach. Turn your ideas into codebundles quickly, and tune them later!

Command and conquer - add CLI commands to your codebundles with a single click.

Supercharge your productivity, simulate commands output and test your regexes.

Instant boilerplate generation

Use presets to prepare your workspace and focus on what delivers the actual value.

Use natural language to describe issues

Simply describe conditions when the issue should be raised. Review and submit the response.

Test generated codebundles in the browser

Local codecollection development environment integration. Try it, run dev environment from this branch generator api follow readme instructions and click on ▶️ button in Codebundle Generator web UI.

Make AI guess your regex

Click on guess regex button and enjoy the magic.

Compose your own regexes

Use natural language to describe the regex you need.

Adjust simulated output

Impact the simulated output by describing the expected content.

Functionalities

Tasks

• User can operate within a single task

  Commands

• Adding command
• Simulating command output
• Parsing command with a regex
• Listing and removing commands from the workspace
• Guessing a regex based on the simulated command output
• Generating regex based on natural language description

Variables/Secrets

• Creating and deleting variables and secrets

Raising Issues

• Describing raise issue conditions in a natural language
• Using form to generate an issue
• Raising issues based on the previously defined commands results
• Stdout parser support

Devtools integration

• running generated robot files in the UI

  Coming soon

• More LLM-based features

  Installation guide

  Clone repository

  git clone git@github.com:Peikae3IWaeG/robotgen.git
  cd robotgen

  Docker

  Create .env file and set variables

  cp .env.example .env

  In order to use Codebundle Generator you must have an OpenAI account with an API key generated. In case of any problems please refer to OpenAI documentation.

  Build and run containers

  docker compose up

  Docker compose setup supports hot code reload.

Local

Requirements

• Python 3.10 (virtual environment recommended)
• NodeJS 18

  API

  cd back
  pip install -r requirements.txt
  OPENAI_API_KEY="sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx" DEV_API_URL="http://dev-api:8001" python main.py

  Frontend

  cd front
  npm install 
  npm run dev

Devtools integration

Codebundle generator needs network access to devtools container. The url is configured with DEV_API_URL and is set to http://dev-api:8001 by default. Following steps are required to use devtools integration feature:

• Clone this repository and switch to generator-api-m1 branch.
• Build the docker image.
• Open your codecollection repository, replace Dockerfile FROM image with the one built in the previous step, and build the image.
• Run the devlopment environment using following command docker run --name dev-api --network robotgen_default --rm -p 8001:8001 -v $PWD:/app/codecollection -v $PWD/auth:/app/auth --env-file .env REPLACE_ME. Container name matches the devault value of DEV_API_URL environment variable. robotgen_default network matches network created by running compose file from this repository and provides network access for codebundle generator API to codecollection development environment. Replace REPLACE_ME with tag/sha of the image built in codecollection repository (previous step)

There's no progres bar/spinner when running a codebundle in the WebUI. If unsure, investigate the development environment container logs.

Open application

UI http://127.0.0.1:3127

API http://127.0.0.1:5127 (swagger generated web ui api docs)

Features suggestions

Feel free to open a discussion and tell me about features you'd like to see.

License

This software is distributed under Apache 2.0 License.