Installing GPT for Confluence:

• Log into your Confluence instance as administrator.
• Click the Apps dropdown and choose Find new apps. The Find new apps screen loads.
• Search for GPT for Confluence and click on the app tile.The App Details screen loads.
• Click Try it free to start installing your app.
• You're all set!Click Close in the Installed and ready to go dialog.

Using GPT for Confluence as a user:

• Step 1: Once the installation is complete, go to the Confluence page where you want to generate text.
• Step 2: Click on the "Aide" link at the top of your article (byline).
• Step 3: A dialog box will appear, prompting you to enter some details about the text you want to generate.
• Step 4: Enter the necessary details, such as the type of content you want to generate, the language, and any other relevant information.
• Step 5: Click "Generate", and the text will appear on the page.
• Step 6: Review the generated text and make any necessary edits or adjustments.
• Step 7: Copy the text as needed as use it anywhere in Confluence.

Tips for using GPT for Confluence:

• Keep your requests as specific as possible to get the most accurate results.
• Don't rely solely on generated text - always review and edit to ensure it meets your needs.
• Experiment with different types of content and formats to fully take advantage of the tool's flexibility.

Develop:

• VSCode (recommended)
• node v16.*
• Set up Ngrok: use VSCode extendsion or ngrok command to setup your own API key
• Crends Credentials.json based on Credentials.json.sample. api-token Reference
• Start pg db: make pg. It'll init db automantically.
• Set local environment:

  export DATABASE_URL=postgres://postgres:postgres@127.0.0.1:5438/gpt_ace
  export LOCAL_BASE_URL=https://{your custom domain}
  export PORT=3000
  export NODE_ENV=production
  export OPENAI_API_KEY=....
  export CF_R2_ACCESS_KEY=....
  export CF_R2_SECRET_KEY=....

• copy .env.sample to .env

• npm install

• npm run build

• npm run start

• install your own app into your Confluence instance at the first time of setting up the local env.

Set up local development environment using Cloudflare Tunnel

With ngrok based local environment setup, one known issue is Failed to start ngrok tunnel on hot reloads. To address this, one alternative is Cloudflare Tunnel.

• Follow this guide to set up the tunnel

• Set up local environment variable for ACE local base url using your public hostname

  e.g.

  export AC_LOCAL_BASE_URL=https://yanhui3000.zenuml.com

• After that, start local server with npm run watch
• Verify the atlassian descriptor at your public hostname, and ensure the baseUrl is correct

Local development of confluence-plugin module

cd modules/confluence-plugin
yarn start:local

Reference

• Getting set up with Atlassian Connect Express (ACE) : https://developer.atlassian.com/cloud/confluence/getting-set-up-with-ace/
• atlassian-connect-express source code: https://bitbucket.org/atlassian/atlassian-connect-express/src/master/

DB ORM

• Sequelize: https://sequelize.org/docs/v6/getting-started/
• Tables:
  • gpt_dock_client_data
  • client_id
  • product_id
  • max_quota
  • token_quota
  • gpt_dock_user_data
  • user_id
  • org_id
  • token_used