Add proactive usage guidance for resource tool

- Clarify that LLM should use this tool proactively
- Add guidance about using before concluding info is unavailable
- Emphasize using instead of suggesting user search themselves
- Should help LLM actually use the tool when needed

Author: kennvene-aws

crates/chat-cli/src/cli/chat/tools/tool_index.json

@@ -234,7 +234,7 @@
   },
   "resource": {
     "name": "resource",
-    "description": "A tool for finding information from the user's indexed knowledge base containing previous conversations, documentation, code, and other materials.\n\n## Overview\nThis tool performs semantic search across indexed content to help answer questions that might have been discussed before or documented in the user's knowledge base.\n\n## When to use\n- When you need information that might be in previous conversations\n- When looking for codebase or project-specific knowledge\n- When the user asks about technical documentation or processes\n- When exploring domain-specific concepts that may have been indexed\n- Before concluding that information is unavailable\n\n## When not to use\n- For general web searches or external information\n- When the knowledge base has not been populated yet\n- For real-time data that changes frequently\n\n## Notes\n- This tool searches across all indexed content using semantic matching\n- For complex questions, break down the query into core concepts to improve search relevance\n- The user's exact phrasing often contains critical contextual cues that enhance semantic matching\n- Try multiple related search terms if the first search doesn't yield relevant results\n- Search results include content from conversations, documentation, and other indexed materials",
+    "description": "A tool for finding information from the user's indexed knowledge base containing previous conversations, documentation, code, and other materials.\n\n## Overview\nThis tool performs semantic search across indexed content to help answer questions that might have been discussed before or documented in the user's knowledge base. Use this tool proactively when you encounter questions you cannot answer from your training data.\n\n## When to use\n- When you need information that might be in previous conversations\n- When looking for codebase or project-specific knowledge\n- When the user asks about technical documentation or processes\n- When exploring domain-specific concepts that may have been indexed\n- Before concluding that information is unavailable or suggesting the user search themselves\n- When the user asks about past discussions or references previous conversations\n\n## When not to use\n- For general web searches or external information\n- When the knowledge base has not been populated yet\n- For real-time data that changes frequently\n\n## Notes\n- This tool searches across all indexed content using semantic matching\n- For complex questions, break down the query into core concepts to improve search relevance\n- The user's exact phrasing often contains critical contextual cues that enhance semantic matching\n- Try multiple related search terms if the first search doesn't yield relevant results\n- Search results include content from conversations, documentation, and other indexed materials",
     "input_schema": {
       "type": "object",
       "properties": {