diff --git a/action.yml b/action.yml
index ee16f514..99459ca1 100644
--- a/action.yml
+++ b/action.yml
@@ -161,61 +161,63 @@ inputs:
       $summary
       ```
 
-      Here is the content of file `$filename` -
+      Here is the content of file `$filename` with 
+      line numbers -
       ```
       $file_content
       ```
 
-      Format for changes and review comments (if any) 
-      on line ranges (line numbers in new file) -
-      <start_line-end_line>: 
-      ```diff
-      <diff_hunk>
+      Format for changes and review comments (if any) -
+      ```new_hunk
+      <new code with line numbers>
       ```
-      ```text
-      <existing review comments>
+
+      ```old_hunk
+      <old code>
       ```
-      ---
-      <start_line-end_line>: 
-      ```diff
-      <diff_hunk>
+
+      ```comment_chain
+      <existing review comments>
       ```
       ---
       ...
 
-      The <diff_hunk> is in the unidiff format.
-
       Changes for review -
       $patches
 
       Your review must consist of comments in the below format 
       with a separator between review comments. The format consists of 
-      line ranges and review comments applicable for that line range. 
-      Any other commentary outside of this format will be ignored 
+      line number ranges and review comments applicable for that line 
+      number range. Any other commentary outside of this format will be ignored 
       and will not be read by the parser -
-      <review_start_line-review_end_line>:
+      <start_line_number_in_new_hunk>-<end_line_number_in_new_hunk>:
       <your review>
       ---
       ...
 
-      Make sure <review_start_line> and <review_end_line> for each review 
-      must be within line ranges in the changes above. Don't echo back the 
+      Example:
+      1-5:
+      LGTM!
+      ---
+
+      It's important that <start_line_number_in_new_hunk> and 
+      <end_line_number_in_new_hunk> for each review must be within 
+      line number ranges of new_hunks. Don't echo back the 
       code provided to you as the line number range is sufficient to map 
-      your comment to the relevant code section in GitHub. Your responses 
-      will be recorded as multi-line review comments on the GitHub pull 
-      request. Markdown format is preferred for text and fenced
-      code blocks should be used for code snippets.
+      your comment to the relevant code section (within new_hunks) 
+      in GitHub. Your responses will be recorded as multi-line review 
+      comments on the GitHub pull request. Markdown format is preferred 
+      for text and fenced code blocks should be used for code snippets.
 
-      Provide thorough and thoughtful feedback to identify any  
-      bug risks or provide improvement suggestions in these diff hunks. 
+      Reflect on the provided code at least 3 times and identify any  
+      bug risks or provide improvement suggestions in these changes. 
       You will point out potential issues such as security, logic errors, 
       syntax errors, out of bound errors, data races, consistency, 
       complexity, error handling, typos, grammar, maintainability, 
       performance, and so on. 
 
-      If there are no issues or suggestions and the diff hunk is acceptable 
-      as-is, please include "LGTM!" (exact word) in your short 
-      review comment.
+      If there are no issues or suggestions and the change is acceptable 
+      as-is, you must include the word `LGTM!` and make a short review comment.
 
   comment:
     required: false
diff --git a/dist/index.js b/dist/index.js
index 1ccdf150..f0be75b4 100644
--- a/dist/index.js
+++ b/dist/index.js
@@ -2438,7 +2438,7 @@ ${tag}`;
                 }
             }
             if (!found) {
-                _actions_core__WEBPACK_IMPORTED_MODULE_0__.info(`Creating new review comment for ${path}:${end_line}: ${message}`);
+                _actions_core__WEBPACK_IMPORTED_MODULE_0__.info(`Creating new review comment for ${path}:${start_line}-${end_line}: ${message}`);
                 await octokit.pulls.createReviewComment({
                     owner: repo.owner,
                     repo: repo.repo,
@@ -2453,7 +2453,7 @@ ${tag}`;
             }
         }
         catch (e) {
-            _actions_core__WEBPACK_IMPORTED_MODULE_0__.warning(`Failed to post review comment: ${e}`);
+            _actions_core__WEBPACK_IMPORTED_MODULE_0__.warning(`Failed to post review comment, for ${path}:${start_line}-${end_line}: ${e}`);
         }
     }
     async review_comment_reply(pull_number, top_level_comment, message) {
@@ -2507,9 +2507,11 @@ ${COMMENT_REPLY_TAG}
     async get_comments_at_range(pull_number, path, start_line, end_line) {
         const comments = await this.list_review_comments(pull_number);
         return comments.filter((comment) => comment.path === path &&
-            comment.start_line === start_line &&
-            comment.line === end_line &&
-            comment.body !== '');
+            comment.body !== '' &&
+            ((comment.start_line &&
+                comment.start_line >= start_line &&
+                comment.line <= end_line) ||
+                comment.line === end_line));
     }
     async get_conversation_chains_at_range(pull_number, path, start_line, end_line, tag = '') {
         const existing_comments = await this.get_comments_at_range(pull_number, path, start_line, end_line);
@@ -4901,7 +4903,7 @@ const codeReview = async (lightBot, heavyBot, options, prompts) => {
             filter_selected_files.push(file);
         }
     }
-    // find patches to review
+    // find hunks to review
     const filtered_files_to_review = await Promise.all(filter_selected_files.map(async (file) => {
         // retrieve file contents
         let file_content = '';
@@ -4934,12 +4936,27 @@ const codeReview = async (lightBot, heavyBot, options, prompts) => {
         const patches = [];
         for (const patch of split_patch(file.patch)) {
             const patch_lines = patch_start_end_line(patch);
-            if (!patch_lines ||
-                patch_lines.start_line === -1 ||
-                patch_lines.end_line === -1) {
+            if (!patch_lines) {
                 continue;
             }
-            patches.push([patch_lines.start_line, patch_lines.end_line, patch]);
+            const hunks = parse_patch(patch);
+            if (!hunks) {
+                continue;
+            }
+            const hunks_str = `
+\`\`\`new_hunk
+${hunks.new_hunk}
+\`\`\`
+
+\`\`\`old_hunk
+${hunks.old_hunk}
+\`\`\`
+`;
+            patches.push([
+                patch_lines.new_hunk.start_line,
+                patch_lines.new_hunk.end_line,
+                hunks_str
+            ]);
         }
         if (patches.length > 0) {
             return [file.filename, file_content, file_diff, patches];
@@ -5098,6 +5115,13 @@ ${skipped_files_to_summarize.length > 0
                     else {
                         comment_chain = '';
                     }
+                    // check comment_chain tokens and skip if too long
+                    const comment_chain_tokens = tokenizer/* get_token_count */.u(comment_chain);
+                    if (comment_chain_tokens >
+                        options.heavy_token_limits.extra_content_tokens) {
+                        core.info(`skip sending comment chain of file: ${ins.filename} due to token count: ${comment_chain_tokens}`);
+                        comment_chain = '';
+                    }
                 }
                 catch (e) {
                     if (e instanceof build/* ChatGPTError */.sK) {
@@ -5105,14 +5129,11 @@ ${skipped_files_to_summarize.length > 0
                     }
                 }
                 ins.patches += `
-${start_line}-${end_line}:
-\`\`\`diff
 ${patch}
-\`\`\`
 `;
                 if (comment_chain !== '') {
                     ins.patches += `
-\`\`\`text
+\`\`\`comment_chain
 ${comment_chain}
 \`\`\`
 `;
@@ -5208,17 +5229,60 @@ const patch_start_end_line = (patch) => {
     const pattern = /(^@@ -(\d+),(\d+) \+(\d+),(\d+) @@)/gm;
     const match = pattern.exec(patch);
     if (match) {
-        const begin = parseInt(match[4]);
-        const diff = parseInt(match[5]);
+        const old_begin = parseInt(match[2]);
+        const old_diff = parseInt(match[3]);
+        const new_begin = parseInt(match[4]);
+        const new_diff = parseInt(match[5]);
         return {
-            start_line: begin,
-            end_line: begin + diff - 1
+            old_hunk: {
+                start_line: old_begin,
+                end_line: old_begin + old_diff - 1
+            },
+            new_hunk: {
+                start_line: new_begin,
+                end_line: new_begin + new_diff - 1
+            }
         };
     }
     else {
         return null;
     }
 };
+const parse_patch = (patch) => {
+    const hunkInfo = patch_start_end_line(patch);
+    if (!hunkInfo) {
+        return null;
+    }
+    const old_hunk_lines = [];
+    const new_hunk_lines = [];
+    //let old_line = hunkInfo.old_hunk.start_line
+    let new_line = hunkInfo.new_hunk.start_line;
+    const lines = patch.split('\n').slice(1); // Skip the @@ line
+    // Remove the last line if it's empty
+    if (lines[lines.length - 1] === '') {
+        lines.pop();
+    }
+    for (const line of lines) {
+        if (line.startsWith('-')) {
+            old_hunk_lines.push(`${line.substring(1)}`);
+            //old_line++
+        }
+        else if (line.startsWith('+')) {
+            new_hunk_lines.push(`${new_line}: ${line.substring(1)}`);
+            new_line++;
+        }
+        else {
+            old_hunk_lines.push(`${line}`);
+            new_hunk_lines.push(`${new_line}: ${line}`);
+            //old_line++
+            new_line++;
+        }
+    }
+    return {
+        old_hunk: old_hunk_lines.join('\n'),
+        new_hunk: new_hunk_lines.join('\n')
+    };
+};
 function parseOpenAIReview(response, debug = false) {
     const reviews = new Map();
     // Split the response into lines
diff --git a/src/commenter.ts b/src/commenter.ts
index 7575b9da..4793d5b3 100644
--- a/src/commenter.ts
+++ b/src/commenter.ts
@@ -171,7 +171,7 @@ ${tag}`
 
       if (!found) {
         core.info(
-          `Creating new review comment for ${path}:${end_line}: ${message}`
+          `Creating new review comment for ${path}:${start_line}-${end_line}: ${message}`
         )
         await octokit.pulls.createReviewComment({
           owner: repo.owner,
@@ -186,7 +186,9 @@ ${tag}`
         })
       }
     } catch (e) {
-      core.warning(`Failed to post review comment: ${e}`)
+      core.warning(
+        `Failed to post review comment, for ${path}:${start_line}-${end_line}: ${e}`
+      )
     }
   }
 
@@ -253,9 +255,11 @@ ${COMMENT_REPLY_TAG}
     return comments.filter(
       (comment: any) =>
         comment.path === path &&
-        comment.start_line === start_line &&
-        comment.line === end_line &&
-        comment.body !== ''
+        comment.body !== '' &&
+        ((comment.start_line &&
+          comment.start_line >= start_line &&
+          comment.line <= end_line) ||
+          comment.line === end_line)
     )
   }
 
diff --git a/src/review.ts b/src/review.ts
index 7f121892..b207fa2a 100644
--- a/src/review.ts
+++ b/src/review.ts
@@ -79,7 +79,7 @@ export const codeReview = async (
     }
   }
 
-  // find patches to review
+  // find hunks to review
   const filtered_files_to_review: (
     | [string, string, string, [number, number, string][]]
     | null
@@ -120,14 +120,27 @@ export const codeReview = async (
       const patches: [number, number, string][] = []
       for (const patch of split_patch(file.patch)) {
         const patch_lines = patch_start_end_line(patch)
-        if (
-          !patch_lines ||
-          patch_lines.start_line === -1 ||
-          patch_lines.end_line === -1
-        ) {
+        if (!patch_lines) {
+          continue
+        }
+        const hunks = parse_patch(patch)
+        if (!hunks) {
           continue
         }
-        patches.push([patch_lines.start_line, patch_lines.end_line, patch])
+        const hunks_str = `
+\`\`\`new_hunk
+${hunks.new_hunk}
+\`\`\`
+
+\`\`\`old_hunk
+${hunks.old_hunk}
+\`\`\`
+`
+        patches.push([
+          patch_lines.new_hunk.start_line,
+          patch_lines.new_hunk.end_line,
+          hunks_str
+        ])
       }
       if (patches.length > 0) {
         return [file.filename, file_content, file_diff, patches]
@@ -348,6 +361,17 @@ ${
           } else {
             comment_chain = ''
           }
+          // check comment_chain tokens and skip if too long
+          const comment_chain_tokens = tokenizer.get_token_count(comment_chain)
+          if (
+            comment_chain_tokens >
+            options.heavy_token_limits.extra_content_tokens
+          ) {
+            core.info(
+              `skip sending comment chain of file: ${ins.filename} due to token count: ${comment_chain_tokens}`
+            )
+            comment_chain = ''
+          }
         } catch (e: unknown) {
           if (e instanceof ChatGPTError) {
             core.warning(
@@ -356,14 +380,11 @@ ${
           }
         }
         ins.patches += `
-${start_line}-${end_line}:
-\`\`\`diff
 ${patch}
-\`\`\`
 `
         if (comment_chain !== '') {
           ins.patches += `
-\`\`\`text
+\`\`\`comment_chain
 ${comment_chain}
 \`\`\`
 `
@@ -484,21 +505,74 @@ const split_patch = (patch: string | null | undefined): string[] => {
 
 const patch_start_end_line = (
   patch: string
-): {start_line: number; end_line: number} | null => {
+): {
+  old_hunk: {start_line: number; end_line: number}
+  new_hunk: {start_line: number; end_line: number}
+} | null => {
   const pattern = /(^@@ -(\d+),(\d+) \+(\d+),(\d+) @@)/gm
   const match = pattern.exec(patch)
   if (match) {
-    const begin = parseInt(match[4])
-    const diff = parseInt(match[5])
+    const old_begin = parseInt(match[2])
+    const old_diff = parseInt(match[3])
+    const new_begin = parseInt(match[4])
+    const new_diff = parseInt(match[5])
     return {
-      start_line: begin,
-      end_line: begin + diff - 1
+      old_hunk: {
+        start_line: old_begin,
+        end_line: old_begin + old_diff - 1
+      },
+      new_hunk: {
+        start_line: new_begin,
+        end_line: new_begin + new_diff - 1
+      }
     }
   } else {
     return null
   }
 }
 
+const parse_patch = (
+  patch: string
+): {old_hunk: string; new_hunk: string} | null => {
+  const hunkInfo = patch_start_end_line(patch)
+  if (!hunkInfo) {
+    return null
+  }
+
+  const old_hunk_lines: string[] = []
+  const new_hunk_lines: string[] = []
+
+  //let old_line = hunkInfo.old_hunk.start_line
+  let new_line = hunkInfo.new_hunk.start_line
+
+  const lines = patch.split('\n').slice(1) // Skip the @@ line
+
+  // Remove the last line if it's empty
+  if (lines[lines.length - 1] === '') {
+    lines.pop()
+  }
+
+  for (const line of lines) {
+    if (line.startsWith('-')) {
+      old_hunk_lines.push(`${line.substring(1)}`)
+      //old_line++
+    } else if (line.startsWith('+')) {
+      new_hunk_lines.push(`${new_line}: ${line.substring(1)}`)
+      new_line++
+    } else {
+      old_hunk_lines.push(`${line}`)
+      new_hunk_lines.push(`${new_line}: ${line}`)
+      //old_line++
+      new_line++
+    }
+  }
+
+  return {
+    old_hunk: old_hunk_lines.join('\n'),
+    new_hunk: new_hunk_lines.join('\n')
+  }
+}
+
 type Review = {
   start_line: number
   end_line: number