diff --git a/README.md b/README.md
index d7acae2..a8d0427 100644
--- a/README.md
+++ b/README.md
@@ -87,8 +87,8 @@ A structured build plan for developing and deploying the AlphaBot Discord compan
 ---
 
 ### 🔥 Day 4 – Roast Command
-- [ ] Build `!roast @user` with AI-generated replies
-- [ ] Style roast output with selected personality
+- [x] Build `!roast @user` with AI-generated replies
+- [x] Style roast output with selected personality
 - [ ] Add cooldown to prevent spam (optional)
 
 ---
diff --git a/bot.py b/bot.py
index fe6abad..b9bb339 100644
--- a/bot.py
+++ b/bot.py
@@ -1,9 +1,26 @@
+# bot.py
+
 import os
 import discord
+import yaml
 from discord.ext import commands
 from dotenv import load_dotenv
 from ai import get_ai_response
 from personality import apply_personality, set_persona
+from discord.ext.commands import (
+    cooldown,
+    BucketType,
+    CooldownMapping,
+    CommandOnCooldown
+)
+import yaml
+
+with open("settings.yml", "r") as f:
+    settings = yaml.safe_load(f)
+
+ROAST_COOLDOWN_SECONDS = settings["cooldowns"]["roast"]
+GLOBAL_COOLDOWN_SECONDS = settings["cooldowns"]["global"]
+COOLDOWN_MSG_TEMPLATE = settings["messages"]["cooldown"]
 
 load_dotenv()
 TOKEN = os.getenv("DISCORD_TOKEN")
@@ -14,8 +31,26 @@ intents.message_content = True
 bot = commands.Bot(command_prefix="!", intents=intents)
 
 @bot.event
-async def on_ready():
-    print(f"✅ Logged in as {bot.user.name}")
+async def on_command_error(ctx, error):
+    if isinstance(error, CommandOnCooldown):
+        retry_secs = round(error.retry_after, 1)
+        msg = COOLDOWN_MSG_TEMPLATE.replace("{seconds}", str(retry_secs))
+        print("🕒 Chill, mortal. You must wait 11.6s before trying again. 😼")
+        await ctx.send(msg)
+    else:
+        raise error
+
+# Global cooldown bucket
+global_cooldown = CooldownMapping.from_cooldown(1, GLOBAL_COOLDOWN_SECONDS, BucketType.user)
+
+@bot.check
+async def global_command_cooldown(ctx):
+    bucket = global_cooldown.get_bucket(ctx.message)
+    retry_after = bucket.update_rate_limit()
+
+    if retry_after:
+        raise CommandOnCooldown(bucket, retry_after, BucketType.user)
+    return True
 
 @bot.command()
 async def ping(ctx):
@@ -33,5 +68,23 @@ async def setpersona(ctx, *, description):
     await ctx.send("✅ Persona updated! New style will be used in replies.")
 
 
+@bot.command(name='roast')
+@cooldown(rate=1, per=ROAST_COOLDOWN_SECONDS, type=BucketType.user)
+async def roast(ctx):
+    # Get the mentioned user (or fallback to the author)
+    target = ctx.message.mentions[0].mention if ctx.message.mentions else ctx.author.mention
+
+    # Build the roast prompt
+    prompt = f"Roast {target}. Be dramatic, insulting, and sarcastic. Speak in your usual chaotic RGB catgirl personality."
+
+    # Get AI response
+    response = get_ai_response(prompt)
+
+    # Send the roast back to the channel
+    await ctx.send(f"😼 {response}")
+
+@bot.event
+async def on_ready():
+    print(f"✅ Logged in as {bot.user.name}")
 
 bot.run(TOKEN)
diff --git a/conda_env.yml b/conda_env.yml
index 0eacaef..2fd886b 100644
--- a/conda_env.yml
+++ b/conda_env.yml
@@ -8,3 +8,4 @@ dependencies:
   - pip:
       - discord.py
       - python-dotenv
+      - pyyaml
diff --git a/requirements.txt b/requirements.txt
index c3104a8..b19c107 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,2 +1,3 @@
 discord.py 
-python-dotenv
\ No newline at end of file
+python-dotenv
+pyyaml
\ No newline at end of file
diff --git a/settings.yml b/settings.yml
new file mode 100644
index 0000000..224a929
--- /dev/null
+++ b/settings.yml
@@ -0,0 +1,5 @@
+cooldowns:
+  global: 15  # seconds between *any* commands from the same user
+  roast: 60  # seconds
+messages:
+  cooldown: "🕒 Chill, mortal. You must wait {seconds}s before trying again. 😼"
\ No newline at end of file