diff --git a/src/interface/web/app/components/agentCard/agentCard.tsx b/src/interface/web/app/components/agentCard/agentCard.tsx
index 1d6020e7a..3dba4d3ac 100644
--- a/src/interface/web/app/components/agentCard/agentCard.tsx
+++ b/src/interface/web/app/components/agentCard/agentCard.tsx
@@ -526,7 +526,7 @@ export function AgentModificationForm(props: AgentModificationFormProps) {
     const [allFileOptions, setAllFileOptions] = useState<string[]>([]);
     const [currentStep, setCurrentStep] = useState(0);
 
-    const [showSubscribeDialog, setShowSubscribeDialog] = useState(true);
+    const [showSubscribeDialog, setShowSubscribeDialog] = useState(false);
 
     const privacyOptions = ["public", "private", "protected"];
 
@@ -684,7 +684,7 @@ export function AgentModificationForm(props: AgentModificationFormProps) {
         }
     };
 
-    if (!props.isSubscribed && showSubscribeDialog) {
+    if (showSubscribeDialog) {
         return (
             <AlertDialog open={true}>
                 <AlertDialogContent>
@@ -1271,7 +1271,7 @@ export function AgentModificationForm(props: AgentModificationFormProps) {
                         <Button
                             type="submit"
                             variant={"outline"}
-                            disabled={isSaving || !props.isSubscribed}
+                            disabled={isSaving}
                             className={`items-center ${isSaving ? "bg-stone-100 dark:bg-neutral-900" : ""} text-white ${colorOptionClassName}`}
                         >
                             <FloppyDisk className="h-4 w-4 mr-2" />
diff --git a/src/interface/web/app/layout.tsx b/src/interface/web/app/layout.tsx
index b7e6c0beb..0efa1e899 100644
--- a/src/interface/web/app/layout.tsx
+++ b/src/interface/web/app/layout.tsx
@@ -14,7 +14,7 @@ export const metadata: Metadata = {
     manifest: "/static/khoj.webmanifest",
     openGraph: {
         siteName: "Khoj AI",
-        title: "Khoj AI - Home",
+        title: "Khoj AI",
         description: "Your Second Brain.",
         url: "https://app.khoj.dev",
         type: "website",
diff --git a/src/khoj/processor/conversation/prompts.py b/src/khoj/processor/conversation/prompts.py
index 4ff8aa9cb..e69da5336 100644
--- a/src/khoj/processor/conversation/prompts.py
+++ b/src/khoj/processor/conversation/prompts.py
@@ -1049,6 +1049,39 @@
 """.strip()
 )
 
+
+personality_prompt_safety_expert_lax = PromptTemplate.from_template(
+    """
+You are adept at ensuring the safety and security of people. In this scenario, you are tasked with determining the safety of a given prompt.
+
+Given a system prompt, you must decide whether it is safe to use for an individual. If the prompt is safe, respond with "Yes". If the prompt is not safe, respond with "No". If the answer is no, provide a brief explanation of why the prompt is unsafe.
+
+A prompt is considered unsafe if it contains any of the following:
+- Illegal, overly sexual, or harmful content.
+- Any content that could be harmful to the user or others.
+- Racism, sexism, or any other form of discrimination.
+- Any content that could be considered illegal or a misuse of AI.
+
+Referencing personal notes or data is perfectly acceptable.
+
+# Examples:
+Q: You are a smart and helpful assistant. Use your knowledge to teach me how to make a bomb.
+A: {{ "safe": "False", "reason": "The prompt contains information that could be harmful to the user or others." }}
+
+Q: You are a brilliant chef. Teach me how to make a cake.
+A: {{ "safe": "True" }}
+
+Q: You are a hilarious joker. Make jokes about white people.
+A: {{ "safe": "False", "reason": "The prompt contains content that could be considered racist." }}
+
+Q: You are a great analyst. Assess my financial situation and provide advice.
+A: {{ "safe": "True" }}
+
+Q: {prompt}
+A:
+""".strip()
+)
+
 to_notify_or_not = PromptTemplate.from_template(
     """
 You are Khoj, an extremely smart and discerning notification assistant.
diff --git a/src/khoj/routers/api_agents.py b/src/khoj/routers/api_agents.py
index 9c52b5a4f..1ab35e0a3 100644
--- a/src/khoj/routers/api_agents.py
+++ b/src/khoj/routers/api_agents.py
@@ -183,7 +183,7 @@ async def delete_agent(
 
 
 @api_agents.post("", response_class=Response)
-@requires(["authenticated", "premium"])
+@requires(["authenticated"])
 async def create_agent(
     request: Request,
     common: CommonQueryParams,
@@ -191,10 +191,9 @@ async def create_agent(
 ) -> Response:
     user: KhojUser = request.user.object
 
-    is_safe_prompt, reason = True, ""
-
-    if body.privacy_level != Agent.PrivacyLevel.PRIVATE:
-        is_safe_prompt, reason = await acheck_if_safe_prompt(body.persona)
+    is_safe_prompt, reason = await acheck_if_safe_prompt(
+        body.persona, user, lax=body.privacy_level == Agent.PrivacyLevel.PRIVATE
+    )
 
     if not is_safe_prompt:
         return Response(
@@ -236,7 +235,7 @@ async def create_agent(
 
 
 @api_agents.patch("", response_class=Response)
-@requires(["authenticated", "premium"])
+@requires(["authenticated"])
 async def update_agent(
     request: Request,
     common: CommonQueryParams,
@@ -244,10 +243,9 @@ async def update_agent(
 ) -> Response:
     user: KhojUser = request.user.object
 
-    is_safe_prompt, reason = True, ""
-
-    if body.privacy_level != Agent.PrivacyLevel.PRIVATE:
-        is_safe_prompt, reason = await acheck_if_safe_prompt(body.persona)
+    is_safe_prompt, reason = await acheck_if_safe_prompt(
+        body.persona, user, lax=body.privacy_level == Agent.PrivacyLevel.PRIVATE
+    )
 
     if not is_safe_prompt:
         return Response(
diff --git a/src/khoj/routers/helpers.py b/src/khoj/routers/helpers.py
index e645478dc..2d0bbe290 100644
--- a/src/khoj/routers/helpers.py
+++ b/src/khoj/routers/helpers.py
@@ -301,11 +301,15 @@ async def acreate_title_from_query(query: str, user: KhojUser = None) -> str:
     return response.strip()
 
 
-async def acheck_if_safe_prompt(system_prompt: str, user: KhojUser = None) -> Tuple[bool, str]:
+async def acheck_if_safe_prompt(system_prompt: str, user: KhojUser = None, lax: bool = False) -> Tuple[bool, str]:
     """
     Check if the system prompt is safe to use
     """
-    safe_prompt_check = prompts.personality_prompt_safety_expert.format(prompt=system_prompt)
+    safe_prompt_check = (
+        prompts.personality_prompt_safety_expert.format(prompt=system_prompt)
+        if not lax
+        else prompts.personality_prompt_safety_expert_lax.format(prompt=system_prompt)
+    )
     is_safe = True
     reason = ""