update prompt to further reduce noise (#201)

<!-- This is an auto-generated comment: release notes by openai -->
### Summary by OpenAI

### Release Notes

- Documentation: Clarified instructions for code review to reduce noise
and improve the review process.

> "Code review made better,  
> With clearer instructions to fetter.  
> Noise reduced, issues highlighted,  
> Reviewers can now be more delighted."
<!-- end of auto-generated comment: release notes by openai -->

Author: harjotgill

action.yml

@@ -102,10 +102,11 @@ inputs:
         - Complexity
         - Optimization
 
-      Refrain from addressing minor code style issues, compliments, or missing 
-      comments/documentation, unless explicitly requested. Concentrate on 
-      identifying and resolving significant concerns to improve overall code 
-      quality while deliberately disregarding minor issues.
+      Refrain from commenting on minor code style issues, missing 
+      comments/documentation, or giving compliments, unless explicitly 
+      requested. Concentrate on identifying and resolving significant 
+      concerns to improve overall code quality while deliberately 
+      disregarding minor issues.
   summarize_file_diff:
     required: false
     description: 'The prompt for file'

dist/index.js

@@ -412,7 +412,8 @@ Important instructions:
   <end_line_number>. The line number range is sufficient to map your 
   comment to the code changes in GitHub pull request.
 - Consider the context provided by the old hunk and associated comment 
-  chain when reviewing the new hunk.
+  chain when reviewing the new hunk. Do not summarize the changes in the 
+  review comment and only focus on pointing out substantive issues.
 - Use Markdown format for review comment text.
 - Fenced code blocks must be used for new content and replacement 
   code/text snippets and must not be annotated with line numbers.
@@ -432,9 +433,9 @@ Important instructions:
   fenced code blocks. These snippets may be added to a different file, such 
   as test cases. Multiple new code snippets are allowed within a single 
   review section.
-- If there are no substantive issues detected at a line range, simply 
-  comment "LGTM!" for the respective line range in a review section and 
-  avoid additional commentary/compliments.
+- If there are no substantive issues detected at a line range and/or the 
+  implementation looks good, you must respond with the comment "LGTM!" and 
+  nothing else for the respective line range in a review section.
 - Reflect on your comments and line number ranges before sending the final 
   response to ensure accuracy of line number ranges and replacement
   snippets.