async function analyzeProfileContent()

in middleware/utils/profileAnalyzer.js [9:50]

• 33 lines of code
• 2 McCabe index (conditional complexity)

async function analyzeProfileContent(markdownContent) {
  try {
    const apiKey = getEnv("GOOGLE_API_KEY")
    const genAI = new GoogleGenerativeAI(apiKey);
    
    // Create a prompt for the LLM
    const prompt = `
You are an expert performance engineer. Please analyze this CPU profile data and provide insights:
1. What are the main performance bottlenecks?
2. Which functions or areas of code need optimization?
3. What specific recommendations would you make to improve performance?

Profile data:
${markdownContent}

Please provide a concise, actionable analysis based on this profile data.
`;
    
    // Get the generative model
    const model = genAI.getGenerativeModel({ model: "gemini-1.5-flash" });
    
    // Make the API call
    const result = await model.generateContent({
      contents: [
        {
          role: "user",
          parts: [{ text: prompt }]
        }
      ],
      generationConfig: {
        temperature: 0.2,
        maxOutputTokens: 2000,
      }
    });

    const response = result.response;
    return response.text();
  } catch (error) {
    console.error("Error analyzing profile:", error);
    throw new Error(`Failed to analyze profile: ${error.message}`);
  }
}