feat: add rate limiter and reduce unnecessary AI calls for free tier

Author: Connor200024

src/echo-ai/aiModel.js

@@ -2,6 +2,33 @@ import { generateText } from 'ai'
 import { aiConfig } from '#configs/ai.config.js'
 import { promptService } from '#ai/services/prompt.service.js'
 
+// Simple rate limiter for OpenAI requests
+class RateLimiter {
+    constructor(maxRequests = 3, windowMs = 60000) {
+        this.maxRequests = maxRequests
+        this.windowMs = windowMs
+        this.requests = []
+    }
+
+    async waitForSlot() {
+        const now = Date.now()
+        // Remove old requests outside the window
+        this.requests = this.requests.filter(time => now - time < this.windowMs)
+        
+        if (this.requests.length >= this.maxRequests) {
+            const oldestRequest = this.requests[0]
+            const waitTime = this.windowMs - (now - oldestRequest) + 100 // Add 100ms buffer
+            console.log(`Rate limit: waiting ${waitTime}ms before next request`)
+            await new Promise(resolve => setTimeout(resolve, waitTime))
+            return this.waitForSlot() // Recursive call after waiting
+        }
+        
+        this.requests.push(now)
+    }
+}
+
+const rateLimiter = new RateLimiter(3, 60000) // 3 requests per minute
+
 /**
  * AI Model wrapper that provides a consistent interface for agents
  */
@@ -72,6 +99,9 @@ export class AIModel {
             // Set a reasonable default max tokens based on prompt length
             const dynamicMaxTokens = this._calculateOptimalTokens(optimizedPrompt)
 
+            // Wait for rate limit slot before making request
+            await rateLimiter.waitForSlot()
+
             // Create a proper system message that won't be overridden
             const result = await generateText({
                 model: this.provider(options.model || aiConfig.model),

src/echo-ai/services/ai.service.js

@@ -633,7 +633,7 @@ class AIService {
      * @private
      */
     _classifyMessageIntent = async message => {
-        // Quick checks for obvious patterns
+        // Quick checks for obvious patterns (no AI call needed)
         const msg = message.toLowerCase()
 
         // Add research pattern check
@@ -690,7 +690,12 @@ class AIService {
             return 'knowledge'
         }
 
-        // Use AI for more nuanced classification with improved prompt
+        // For simple/short messages, default to conversation without AI call
+        if (message.length < 50 || msg.match(/^(hi|hello|hey|thanks|ok|yes|no|sure)\b/)) {
+            return 'conversation'
+        }
+
+        // Only use AI for complex/ambiguous cases - this reduces API calls significantly
         try {
             const classification = await this.aiModel.getResponse(`Classify this message into one of these categories:
 Message: "${message}"