fix: provide target word to AI while preventing disclosure

- Include actual target word in prompts so AI gives accurate clues
- Add explicit instructions to never mention or reveal the word
- Update system prompt to clarify AI should know word but not disclose it
- Fixes issue where AI was guessing wrong word and giving incorrect clues

src/aiService.js

@@ -44,7 +44,7 @@ class OpenRouterService {
           model: this.model,
           messages: [{
             role: 'system',
-            content: 'You provide clues for word puzzles. Follow the EXACT format requested. Be concise and direct. Never reveal the actual word. Use plain text only - no bold, italics, asterisks, or markdown formatting. Stick to word limits.'
+            content: 'You provide clues for word puzzles. You will be told the target word that players need to guess, but you must NEVER mention, spell, or reveal that word in your response. Follow the EXACT format requested. Be concise and direct about the target word without revealing it. Use plain text only - no bold, italics, asterisks, or markdown formatting. Stick to word limits.'
           }, {
             role: 'user',
             content: prompt

src/conversationManager.js

@@ -74,14 +74,15 @@ class ChatService {
     const bookTitle = context.bookTitle;
     const author = context.author;
     
-    // Create sentence with blank instead of revealing the word
+    // Create sentence with blank for context, but tell AI the actual word
     const sentenceWithBlank = sentence.replace(new RegExp(`\\b${word}\\b`, 'gi'), '____');
     
     try {
-      // Build focused prompt using the conversation manager's logic
+      // Build focused prompt that includes the target word but forbids revealing it
       const prompt = this.buildFocusedPrompt({
         ...context,
-        sentence: sentenceWithBlank
+        sentence: sentenceWithBlank,
+        targetWord: word
       }, questionType, userInput);
       
       // Use the AI service as a simple API wrapper
@@ -100,20 +101,21 @@ class ChatService {
 
   // Build focused prompt for specific question types
   buildFocusedPrompt(context, questionType, userInput) {
-    const { sentence, bookTitle, author } = context;
+    const { sentence, bookTitle, author, targetWord } = context;
     const baseContext = `From "${bookTitle}" by ${author}: "${sentence}"`;
+    const wordInstruction = `The target word is "${targetWord}". NEVER mention or reveal this word in your response.`;
     
     const prompts = {
-      part_of_speech: `${baseContext}\n\nWhat type of word fits in the blank? Say "This is a [noun/verb/adjective/adverb]" then give a grammatical hint about its function. Maximum 20 words. Never reveal the word.`,
+      part_of_speech: `${baseContext}\n\n${wordInstruction}\n\nGive the part of speech for "${targetWord}" and a grammatical hint. Say "This is a [noun/verb/adjective/adverb]" then explain its function. Maximum 20 words. Do not reveal the actual word.`,
       
-      sentence_role: `${baseContext}\n\nWhat meaning role does the blank serve? Does it express action, emotion, description, or relationship? Maximum 20 words.`,
+      sentence_role: `${baseContext}\n\n${wordInstruction}\n\nDescribe the role "${targetWord}" plays in this sentence. Does it express action, emotion, description, or relationship? Maximum 20 words. Do not reveal the actual word.`,
       
-      word_category: `${baseContext}\n\nWhat category does this word belong to? Say "This word describes [general category]" without giving specific examples. Maximum 20 words.`,
+      word_category: `${baseContext}\n\n${wordInstruction}\n\nWhat category does "${targetWord}" belong to? Say "This word describes [general category]" without giving specific examples. Maximum 20 words. Do not reveal the actual word.`,
       
-      synonym: `${baseContext}\n\nGive a conceptual hint. Format: "Think of something that [general description of concept]." Be indirect and conceptual. Maximum 20 words.`
+      synonym: `${baseContext}\n\n${wordInstruction}\n\nGive a conceptual hint about "${targetWord}". Format: "Think of something that [general description]." Be indirect and conceptual. Maximum 20 words. Do not reveal the actual word.`
     };
     
-    return prompts[questionType] || `${baseContext}\n\nProvide a helpful hint about the missing word without revealing it.`;
+    return prompts[questionType] || `${baseContext}\n\n${wordInstruction}\n\nProvide a helpful hint about "${targetWord}" without revealing it.`;
   }
 
   // Simple fallback responses