OpenAI GPT powered markdown docs checker action

This action uses OpenAI's GPT-3.5-turbo model to ask for pointers on ways in which the repositories markdown docs can be improved.

Important

• The action requires an OpenAI API token secret to be set up on the repository, It is also up to you to pay for any API costs incurred through usage.

Warning

• Using on repositories with many markdown files may be expensive.
• Files that are too large (> ~3000 words) may not process or may have their responses from GPT cut off early due to the model that is limited to ~4000 tokens.
• The responses are not guaranteed to be accurate, or even helpful. Do not follow what it says without verifying its responses first
• The responses may not stick to the grammar, spelling, and general inputs.

Inputs

temperature

Optional Must be between 0 and 1. The temperature parameter to use for GPT responses. A lower value makes responses more factual and less 'creative'. Defaulted to 0.75.

spelling

Optional The spelling parameter denotes whether GPT will be prompted to give suggestions on spelling. Defaulted to true

grammar

Optional The grammar parameter denotes whether GPT will be prompted to give suggestions on grammar. Defaulted to true

general

Optional The general parameter denotes whether GPT will be prompted to give general suggestions and improvements. Defaulted to true

Outputs

result

A string containing the GPT responses for each markdown file.

Example usage

name: GPT Docs Feedback
on:
  push:
    branches:
      - main
env:
  OPENAI_API_KEY: ${{secrets.OPENAI_API_KEY}}

jobs:
  docs-feedback:
    runs-on: ubuntu-latest
    steps:
      - name: Check out repository
        uses: actions/checkout@v2

      - name: Set up Node.js
        uses: actions/setup-node@v2
        with:
          node-version: 14

      - name: Install dependencies
        run: npm ci

      - name: Find and process .md files
        id: find-and-process-md
        uses: Hamish-taylor/GPT-Docs-Checker@main
        with:
            temperature: '1'
            spelling: false
            grammar: true
            general: true
               
      - name: Print the result
        run: |
          echo "Markdown-processing result:"
          echo "$RESULT"
        env:
          RESULT: ${{ steps.find-and-process-md.outputs.result }}