OpenAI与JIRA集成POC开发求助：实现从JIRA取数并通过OpenAI分析的最小可运行方案

Hey there! Let's put together a minimal, working POC for this—since you just need it to run once, we can skip over complex frontend setup hurdles and go with a straightforward Node.js script that handles both JIRA data fetching and OpenAI analysis. Here's how to make it happen step by step:

1. Prep Your Credentials #

First, grab the keys you'll need:

• JIRA: Get your API token (go to your JIRA profile > Security > Create and manage API tokens) and note your JIRA email + your JIRA domain (like your-domain.atlassian.net).
• OpenAI: Grab your API key from your OpenAI account settings.

2. Minimal Node.js Script (All-in-One) #

We'll use two npm packages: axios for JIRA API calls, and openai for interacting with OpenAI. Create a new folder, set up the script, and install dependencies:

First, initialize the project and install packages:

npm init -y
npm install axios openai dotenv

Then create a file jira-openai-poc.js with this code:

require('dotenv').config();
const axios = require('axios');
const { OpenAI } = require('openai');

// Initialize OpenAI client
const openai = new OpenAI({
  apiKey: process.env.OPENAI_API_KEY,
});

// JIRA configuration
const jiraConfig = {
  email: process.env.JIRA_EMAIL,
  token: process.env.JIRA_TOKEN,
  domain: process.env.JIRA_DOMAIN,
};

// Function to fetch JIRA issues via JQL
async function fetchJiraIssues(jql) {
  try {
    const response = await axios.get(`https://${jiraConfig.domain}/rest/api/3/search`, {
      auth: {
        username: jiraConfig.email,
        password: jiraConfig.token,
      },
      params: {
        jql: jql,
        fields: 'summary,description,status', // Keep fields minimal for POC
        maxResults: 5, // Limit results to keep it simple
      },
    });

    return response.data.issues.map(issue => ({
      key: issue.key,
      summary: issue.fields.summary,
      description: issue.fields.description || 'No description',
      status: issue.fields.status.name,
    }));
  } catch (error) {
    console.error('Error fetching JIRA issues:', error.response?.data || error.message);
    throw error;
  }
}

// Function to analyze issues with OpenAI
async function analyzeWithOpenAI(issues) {
  try {
    const prompt = `Summarize the following JIRA issues and highlight the most common status and top 2 recurring themes:\n\n${JSON.stringify(issues, null, 2)}`;

    const completion = await openai.chat.completions.create({
      model: 'gpt-3.5-turbo',
      messages: [{ role: 'user', content: prompt }],
    });

    return completion.choices[0].message.content;
  } catch (error) {
    console.error('Error analyzing with OpenAI:', error.message);
    throw error;
  }
}

// Main function to run the POC
async function runPOC() {
  // Use a simple JQL query (adjust this to match your project/sprint if needed)
  const jql = 'project = "YOUR_PROJECT_KEY" AND sprint = "YOUR_SPRINT_NAME"';
  
  console.log('Fetching JIRA issues...');
  const issues = await fetchJiraIssues(jql);
  
  console.log('Fetched issues:', issues);
  
  console.log('\nAnalyzing with OpenAI...');
  const analysis = await analyzeWithOpenAI(issues);
  
  console.log('\nOpenAI Analysis Result:\n', analysis);
}

// Run the POC
runPOC();

3. Set Up Environment Variables #

Create a .env file in your project folder with these values:

JIRA_EMAIL=your-jira-email@example.com
JIRA_TOKEN=your-jira-api-token
JIRA_DOMAIN=your-domain.atlassian.net
OPENAI_API_KEY=your-openai-api-key

4. Run the Script #

Just execute this command in your terminal:

node jira-openai-poc.js

Quick Notes on Your Frontend Script #

If you really want to stick with the frontend form you started, you'll need to handle CORS issues (JIRA doesn't allow direct frontend calls by default). For a quick fix, you could use a proxy server, but the Node.js script above is way faster to get up and running for a minimal POC.

That's it—this will fetch a small set of JIRA issues, send them to OpenAI, and return a simple analysis. Adjust the JQL query or the OpenAI prompt to match what you want to analyze!

备注：内容来源于stack exchange，提问作者Cinthiya Singh