n8n Workflow: Score SDK Documentation Localization Readiness with Azure GPT-4o Mini and Slack Alerts

This n8n workflow automates the process of evaluating SDK documentation for localization readiness using Azure's GPT-4o Mini, and then alerts the relevant team via Slack if the documentation is not ready. It also logs the results in Notion and Google Sheets.

What it does

This workflow performs the following key steps:

1. Manual Trigger: Initiates the workflow upon manual execution, allowing for on-demand checks.
2. Google Sheets (Placeholder): (Currently not configured to read/write specific data, but typically would fetch or log documentation details).
3. AI Agent: Acts as an orchestrator, leveraging an Azure OpenAI Chat Model to analyze SDK documentation.
4. Azure OpenAI Chat Model: Utilizes Azure's GPT-4o Mini to assess the localization readiness of the provided SDK documentation.
5. Code: Processes the output from the AI Agent, extracting the localizationReady status and the reasoning for the assessment.
6. If: Checks the localizationReady status.
   • If localizationReady is false, it proceeds to send a Slack alert.
   • If localizationReady is true, it proceeds to log the status in Notion.
7. Slack: Sends an alert message to a designated Slack channel if the documentation is not localization-ready, including the reasoning from the AI.
8. Notion: Logs the localization readiness status and reasoning into a Notion database, regardless of whether it's ready or not.

Prerequisites/Requirements

To use this workflow, you will need:

• n8n Instance: A running n8n instance.
• Azure OpenAI Account: Access to Azure OpenAI with a deployed GPT-4o Mini model. You'll need an API key and endpoint.
• Slack Account: A Slack workspace and a channel where alerts should be posted. You'll need a Slack API token or webhook.
• Notion Account: A Notion workspace and a database where the localization readiness status will be logged. You'll need a Notion API token and the database ID.
• Google Sheets (Optional): If you intend to use the Google Sheets node for data input or output, you'll need a Google account with access to Google Sheets and the appropriate credentials.

Setup/Usage

1. Import the Workflow:
   • Copy the provided JSON workflow definition.
   • In your n8n instance, go to "Workflows" and click "New".
   • Click the "Import from JSON" button and paste the copied JSON.
2. Configure Credentials:
   • Azure OpenAI Chat Model:
     • Click on the "Azure OpenAI Chat Model" node.
     • Select or create a new "Azure OpenAI Chat API" credential.
     • Provide your Azure OpenAI API Key, Endpoint, and the Deployment Name of your GPT-4o Mini model.
   • Slack:
     • Click on the "Slack" node.
     • Select or create a new "Slack API" credential.
     • Provide your Slack Bot User OAuth Token.
     • Specify the "Channel" where you want to receive alerts (e.g., #dev-alerts or #localization-team).
   • Notion:
     • Click on the "Notion" node.
     • Select or create a new "Notion API" credential.
     • Provide your Notion Integration Token.
     • Configure the "Database ID" to which you want to log the results.
     • Map the properties in the Notion node to the relevant data from the workflow (e.g., localizationReady and reasoning).
   • Google Sheets (if used):
     • Click on the "Google Sheets" node.
     • Select or create a new "Google Sheets API" credential (OAuth2 is recommended).
     • Configure the Spreadsheet ID and Sheet Name if you plan to use this node for data.
3. Configure AI Agent (System Prompt):
   • Click on the "AI Agent" node.
   • Review the "System Prompt" to ensure it aligns with how you want the AI to evaluate localization readiness. This prompt will instruct GPT-4o Mini on what criteria to use.
   • Ensure the "Input Data" for the AI Agent is correctly mapped to the SDK documentation content you want to analyze. (The current workflow does not explicitly define where the SDK documentation comes from, you might add a preceding node like an HTTP Request to fetch documentation, or a manual input).
4. Configure Code Node:
   • The "Code" node is set up to extract localizationReady (boolean) and reasoning (string) from the AI Agent's output. Review and adjust the JavaScript code if the AI Agent's output format changes.
5. Configure If Node:
   • The "If" node checks {{ $json.localizationReady }} for false. No changes are typically needed here unless your logic for triggering alerts changes.
6. Activate the Workflow:
   • Once all credentials and configurations are set, activate the workflow by toggling the "Active" switch in the top right corner of the n8n editor.
7. Execute the Workflow:
   • Click "Execute Workflow" on the "Manual Trigger" node to run the workflow.
   • Observe the execution results and check your Slack channel and Notion database for the output.

This workflow provides a robust foundation for automating localization readiness checks, ensuring that your SDK documentation meets internationalization standards before release.