diff --git a/README.md b/README.md
index f9693cdc..5e1c2759 100644
--- a/README.md
+++ b/README.md
@@ -56,13 +56,13 @@ FAQs, you can refer to the sections below.
 - [Contribute](#contribute)
 - [FAQs](#faqs)
 
-## Professional Version of CodeRabbit
+## CodeRabbit Pro
 
 The professional version of `openai-pr-reviewer` project is now available at
-[coderabbit.ai](http://coderabbit.ai). Building upon our open-source foundation,
-CodeRabbit offers premium features including enhanced context and superior noise
-reduction, dedicated support, and our ongoing commitment to improve code
-reviews.
+[coderabbit.ai](http://coderabbit.ai). Building upon our open source foundation,
+CodeRabbit Pro offers premium features including enhanced context and superior
+noise reduction, dedicated support, and our ongoing commitment to improve code
+reviews. Moreover, CodeRabbit Pro is free for open source projects.
 
 ## Install instructions
 
diff --git a/dist/index.js b/dist/index.js
index 37c8e4e1..a6f80920 100644
--- a/dist/index.js
+++ b/dist/index.js
@@ -3917,8 +3917,7 @@ const COMMENT_REPLY_TAG = '<!-- This is an auto-generated reply by OSS CodeRabbi
 const SUMMARIZE_TAG = '<!-- This is an auto-generated comment: summarize by OSS CodeRabbit -->';
 const IN_PROGRESS_START_TAG = '<!-- This is an auto-generated comment: summarize review in progress by OSS CodeRabbit -->';
 const IN_PROGRESS_END_TAG = '<!-- end of auto-generated comment: summarize review in progress by OSS CodeRabbit -->';
-const DESCRIPTION_START_TAG = `
-<!-- This is an auto-generated comment: release notes by OSS CodeRabbit -->`;
+const DESCRIPTION_START_TAG = '<!-- This is an auto-generated comment: release notes by OSS CodeRabbit -->';
 const DESCRIPTION_END_TAG = '<!-- end of auto-generated comment: release notes by OSS CodeRabbit -->';
 const RAW_SUMMARY_START_TAG = `<!-- This is an auto-generated comment: raw summary by OSS CodeRabbit -->
 <!--
@@ -3977,7 +3976,7 @@ ${tag}`;
     }
     removeContentWithinTags(content, startTag, endTag) {
         const start = content.indexOf(startTag);
-        const end = content.indexOf(endTag);
+        const end = content.lastIndexOf(endTag);
         if (start >= 0 && end >= 0) {
             return content.slice(0, start) + content.slice(end + endTag.length);
         }
@@ -4013,7 +4012,7 @@ ${tag}`;
             }
             const description = this.getDescription(body);
             const messageClean = this.removeContentWithinTags(message, DESCRIPTION_START_TAG, DESCRIPTION_END_TAG);
-            const newDescription = `${description}${DESCRIPTION_START_TAG}\n${messageClean}\n${DESCRIPTION_END_TAG}`;
+            const newDescription = `${description}\n${DESCRIPTION_START_TAG}\n${messageClean}\n${DESCRIPTION_END_TAG}`;
             await _octokit__WEBPACK_IMPORTED_MODULE_2__/* .octokit.pulls.update */ .K.pulls.update({
                 owner: repo.owner,
                 repo: repo.repo,
@@ -6783,63 +6782,26 @@ $description
 $short_summary
 \`\`\`
 
-## Instructions
+## IMPORTANT Instructions
 
-The format for changes provided in the example below consists of 
-multiple change sections, each containing a new hunk (annotated with 
-line numbers), an old hunk, and optionally, existing comment chains. 
-Note that the old hunk code has been replaced by the new hunk. Some 
-lines on the new hunk may be annotated with line numbers.
-
-Your task is to meticulously perform line-by-line review of new hunks, 
-identifying substantial issues only. Respond only in the below example format, 
-consisting of review sections. Each review section must have a line number range 
-and a review comment for that range. Use separator after each review section. 
-Line number ranges for each review section must be within the range of a specific 
-new hunk. Start line number must belong to the same hunk as the end line number.
-Provide the exact line number range (inclusive) for each review comment. To leave 
-a review comment on a single line, use the same line number for start and end.
-
-Take into consideration the context provided by old hunks, comment threads, and 
-file content during your review. Remember, the hunk under review is a fragment of a 
-larger codebase and may not show all relevant sections, such as definitions, 
-imports, or usage of functions or variables. Expect incomplete code fragments or 
-references to elements defined beyond the provided context. Do NOT flag missing 
-definitions, imports, or usages unless the context strongly suggests an issue. 
-Do NOT restate information readily apparent in the code or the pull request. 
-Do NOT provide general feedback, summaries, explanations of changes, or praises 
-for making good additions. Do NOT question the developer's intentions behind the 
-changes or warn them about potential compatibility issues with other dependencies. 
-Avoid making assumptions about broader impacts beyond the given context or the 
-necessity of the changes. Do NOT request the developer to review their changes. 
-Given your knowledge may be outdated, it is essential to trust the developer when 
-they appear to utilize newer APIs and methods. Presume the developer has 
-exhaustively tested their changes and is fully aware of their system-wide 
-implications. Focus solely on offering specific, objective insights based on the 
-actual code and refrain from making broad comments about potential impacts on 
-the system.
-
-Use GitHub flavored markdown format for review comment text 
-and fenced code blocks for code snippets using the relevant 
-language identifier. Do NOT annotate the code snippet with 
-line numbers. The code snippet must be correctly 
-formatted & indented.
-
-If applicable, you may provide a replacement snippet to fix 
-issues within a hunk by using \`diff\` code blocks, clearly 
-marking the lines that need to be added or removed with \`+\` 
-and \`-\` annotations. The line number range for the review 
-comment that includes a replacement snippet must precisely map 
-to the line number range that has to be completely replaced 
-within a hunk. Do NOT use \`suggestion\` code blocks for
-replacement snippets.
+Input: New hunks annotated with line numbers and old hunks (replaced code). Hunks represent incomplete code fragments.
+Additional Context: PR title, description, summaries and comment chains.
+Task: Review new hunks for substantive issues using provided context and respond with comments if necessary.
+Output: Review comments in markdown with exact line number ranges in new hunks. Start and end line numbers must be within the same hunk. For single-line comments, start=end line number. Must use example response format below.
+Use fenced code blocks using the relevant language identifier where applicable.
+Don't annotate code snippets with line numbers. Format and indent code correctly.
+Do not use \`suggestion\` code blocks.
+For fixes, use \`diff\` code blocks, marking changes with \`+\` or \`-\`. The line number range for comments with fix snippets must exactly match the range to replace in the new hunk.
+
+- Do NOT provide general feedback, summaries, explanations of changes, or praises 
+  for making good additions. 
+- Focus solely on offering specific, objective insights based on the 
+  given context and refrain from making broad comments about potential impacts on 
+  the system or question intentions behind the changes.
 
 If there are no issues found on a line range, you MUST respond with the 
 text \`LGTM!\` for that line range in the review section. 
 
-Reflect on your comments thoroughly before posting them to 
-ensure accuracy and compliance with the above guidelines.
-
 ## Example
 
 ### Example changes
@@ -7636,7 +7598,7 @@ ${lib_commenter/* SHORT_SUMMARY_END_TAG */.Zb}
 
 ### CodeRabbit Pro
 
-If you like this project, please support us by purchasing the [Pro version](https://coderabbit.ai). The Pro version has advanced context, superior noise reduction and several proprietary improvements compared to the open source version.
+If you like this project, please support us by purchasing the [Pro version](https://coderabbit.ai). The Pro version has advanced context, superior noise reduction and several proprietary improvements compared to the open source version. Moreover, CodeRabbit Pro is free for open source projects.
 
 </details>
 `;
diff --git a/src/commenter.ts b/src/commenter.ts
index 2c6789c9..463e92cf 100644
--- a/src/commenter.ts
+++ b/src/commenter.ts
@@ -24,8 +24,8 @@ export const IN_PROGRESS_START_TAG =
 export const IN_PROGRESS_END_TAG =
   '<!-- end of auto-generated comment: summarize review in progress by OSS CodeRabbit -->'
 
-export const DESCRIPTION_START_TAG = `
-<!-- This is an auto-generated comment: release notes by OSS CodeRabbit -->`
+export const DESCRIPTION_START_TAG =
+  '<!-- This is an auto-generated comment: release notes by OSS CodeRabbit -->'
 export const DESCRIPTION_END_TAG =
   '<!-- end of auto-generated comment: release notes by OSS CodeRabbit -->'
 
@@ -93,7 +93,7 @@ ${tag}`
 
   removeContentWithinTags(content: string, startTag: string, endTag: string) {
     const start = content.indexOf(startTag)
-    const end = content.indexOf(endTag)
+    const end = content.lastIndexOf(endTag)
     if (start >= 0 && end >= 0) {
       return content.slice(0, start) + content.slice(end + endTag.length)
     }
@@ -155,7 +155,7 @@ ${tag}`
         DESCRIPTION_START_TAG,
         DESCRIPTION_END_TAG
       )
-      const newDescription = `${description}${DESCRIPTION_START_TAG}\n${messageClean}\n${DESCRIPTION_END_TAG}`
+      const newDescription = `${description}\n${DESCRIPTION_START_TAG}\n${messageClean}\n${DESCRIPTION_END_TAG}`
       await octokit.pulls.update({
         owner: repo.owner,
         repo: repo.repo,
diff --git a/src/review.ts b/src/review.ts
index 69112de4..5e7dd9db 100644
--- a/src/review.ts
+++ b/src/review.ts
@@ -465,7 +465,7 @@ ${SHORT_SUMMARY_END_TAG}
 
 ### CodeRabbit Pro
 
-If you like this project, please support us by purchasing the [Pro version](https://coderabbit.ai). The Pro version has advanced context, superior noise reduction and several proprietary improvements compared to the open source version.
+If you like this project, please support us by purchasing the [Pro version](https://coderabbit.ai). The Pro version has advanced context, superior noise reduction and several proprietary improvements compared to the open source version. Moreover, CodeRabbit Pro is free for open source projects.
 
 </details>
 `