Skip to content

README.md

Latest commit

 

History

History
138 lines (93 loc) · 4.65 KB

README.md

File metadata and controls

138 lines (93 loc) · 4.65 KB

Auto startup workspace

python FastAPI action mit


Overview

Startup overview

The workspace allows developers to start up a project easily by creating an issue within your requirements.

Usage

Setup repository

  • When you come up with an idea then want to implement a software for that idea. Let's start by creating a new repository in our workspace. Create repository

  • Create GitHub Access Tokens

    • Go to Settings ⟶ Developer Settings ⟶ Personal access tokens ⟶ Generate new token
    • Select All repositories option in Repository access.
    • Grant Read and Write permission for Contents access.
    • Gerate token and copy to clipboard. Setup repository

  • Assign Secret

    • Grant repo token permission to access the repository.
      • From homepage of the created repository. Go to `Settings ⟶ Secrets and variables ⟶ Actions ⟶ New repository secret.
      • Name of the secrets is ACTION_TOKEN
      • Paste the created GitHub Access Tokens.
      • Add secret. Add secret
    • Add Gemini API key as secret variable.
      • Name of the secrets is GEMINI_API_KEY
      • Paste the created Gemini API key.
      • Add secret. Add secret
    • Add Mando API key as secret variable.
      • Name and value of the secrets is API_KEY = MqQVfJ6Fq1umZnUI7ZuaycciCjxi3gM0

Action Workflow

  • From the homepage of the new repository, switch to Action tab to manually set action through set up a workflow yourself →

Setup action

  • This is a simple version of action script.
name: Generate Code from Issue

on:
  issues:
    types: [opened, reopened, edited]

jobs:
  generate-code:
    runs-on: ubuntu-latest
    steps:
    - name: Checkout repository
      uses: actions/checkout@v4
      with:
        token: ${{ secrets.ACTION_TOKEN }}
        
    - name: HyperAgent Generator
      uses: Auto-Actions/hyperagent-action@v2.0.0
      with:
        gemini-api-key: ${{ secrets.GEMINI_API_KEY }}
        api-key: ${{ secrets.API_KEY }}
        github-token: ${{ secrets.ACTION_TOKEN }}
        output-path: 'generated-code'
        programming-language: 'python'
        branch-name: 'feature/ai-generated'
        create-pull-request: 'true'
        model-name: 'gemini-2.5-pro'
  • Commit the file as an initial commit. For further actions, you have to update this file as a new commit.

Initial issues

Create issue

  • Then you can switch to Action tab to monitor the development process.

Action running

  • When the action running finishes, the initial software is already in branch initial.

Checkout source code

Demo video

Youtube

Dataset

The dataset for evaluating the framework is available on Huggingface.