diff --git a/codeflash/after_aiagents.py b/codeflash/after_aiagents.py
new file mode 100644
index 00000000..31395b60
--- /dev/null
+++ b/codeflash/after_aiagents.py
@@ -0,0 +1,15 @@
+def _estimate_string_tokens(content: str | Sequence[UserContent]) -> int:
+    if not content:
+        return 0
+    if isinstance(content, str):
+        return len(_TOKEN_SPLIT_RE.split(content.strip()))
+    tokens = 0
+    for part in content:
+        if isinstance(part, str):
+            tokens += len(_TOKEN_SPLIT_RE.split(part.strip()))
+        elif isinstance(part, BinaryContent):
+            tokens += len(part.data)
+    return tokens
+
+
+_TOKEN_SPLIT_RE = re.compile(r'[\s",.:]+')
diff --git a/codeflash/after_algo.py b/codeflash/after_algo.py
new file mode 100644
index 00000000..bf506b2c
--- /dev/null
+++ b/codeflash/after_algo.py
@@ -0,0 +1,9 @@
+def postprocess(self, predictions: tuple[np.ndarray, ...], max_detections: int):
+    bboxes, logits = predictions
+    batch_size, num_queries, num_classes = logits.shape
+    logits_sigmoid = self.sigmoid_stable(logits)
+    for batch_idx in range(batch_size):
+        logits_flat = logits_sigmoid[batch_idx].reshape(-1)
+        # Use argpartition for better performance when max_detections is smaller than logits_flat
+        partition_indices = np.argpartition(-logits_flat, max_detections)[:max_detections]
+        sorted_indices = partition_indices[np.argsort(-logits_flat[partition_indices])]
diff --git a/codeflash/after_data.py b/codeflash/after_data.py
new file mode 100644
index 00000000..d80080b2
--- /dev/null
+++ b/codeflash/after_data.py
@@ -0,0 +1,3 @@
+def check_missing_data(data: pd.DataFrame):
+    """Check if there is any missing data in the DataFrame"""
+    return data.isnull().values.any()
diff --git a/codeflash/after_db.py b/codeflash/after_db.py
new file mode 100644
index 00000000..ee8a2d48
--- /dev/null
+++ b/codeflash/after_db.py
@@ -0,0 +1,3 @@
+def get_authors(session):
+    query = session.query(Author).join(Book).distinct(Author.id).order_by(Author.id)
+    return query.all()
diff --git a/codeflash/after_ml.py b/codeflash/after_ml.py
new file mode 100644
index 00000000..010ea725
--- /dev/null
+++ b/codeflash/after_ml.py
@@ -0,0 +1,11 @@
+import jax.numpy as jnp
+
+
+def sum_hand(hand):
+    """Returns the total points in a hand."""
+    return jnp.sum(hand) + (10 * usable_ace(hand))
+
+
+def usable_ace(hand):
+    """Checks to se if a hand has a usable ace."""
+    return jnp.logical_and(jnp.any(hand == 1), jnp.sum(hand) + 10 <= 21)
diff --git a/codeflash/after_numerical.py b/codeflash/after_numerical.py
new file mode 100644
index 00000000..98bfc9f4
--- /dev/null
+++ b/codeflash/after_numerical.py
@@ -0,0 +1,17 @@
+def _equalize_cv(img: np.ndarray, mask: np.ndarray | None = None) -> np.ndarray:
+    histogram = cv2.calcHist([img], [0], mask, [256], (0, 256)).ravel()
+
+    # Find the first non-zero index with a numpy operation
+    i = np.flatnonzero(histogram)[0] if np.any(histogram) else 255
+
+    total = np.sum(histogram)
+    if histogram[i] == total:
+        return np.full_like(img, i)
+
+    scale = 255.0 / (total - histogram[i])
+
+    # Optimize cumulative sum and scale to generate LUT
+    cumsum_histogram = np.cumsum(histogram)
+    lut = np.clip(((cumsum_histogram - cumsum_histogram[i]) * scale)
+                  .round(), 0, 255).astype(np.uint8)
+    return sz_lut(img, lut, inplace=True)
diff --git a/codeflash/after_web_server.py b/codeflash/after_web_server.py
new file mode 100644
index 00000000..8286fb58
--- /dev/null
+++ b/codeflash/after_web_server.py
@@ -0,0 +1,11 @@
+async def get_endpoint(session, url):
+    async with session.get(url) as response:
+        return await response.text()
+
+
+async def some_api_call(urls):
+    async with aiohttp.ClientSession() as session:
+        tasks = [get_endpoint(session, url) for url in urls]
+        # Run requests concurrently
+        results = await asyncio.gather(*tasks)
+        return results
diff --git a/codeflash/before_aiagents.py b/codeflash/before_aiagents.py
new file mode 100644
index 00000000..f5d0f8ea
--- /dev/null
+++ b/codeflash/before_aiagents.py
@@ -0,0 +1,16 @@
+def _estimate_string_tokens(content: str | Sequence[UserContent]) -> int:
+    if not content:
+        return 0
+    if isinstance(content, str):
+        return len(re.split(r'[\s",.:]+', content.strip()))
+    tokens = 0
+    for part in content:
+        if isinstance(part, str):
+            tokens += len(re.split(r'[\s",.:]+', part.strip()))
+        if isinstance(part, (AudioUrl, ImageUrl)):
+            tokens += 0
+        elif isinstance(part, BinaryContent):
+            tokens += len(part.data)
+        else:
+            tokens += 0
+    return tokens
diff --git a/codeflash/before_algo.py b/codeflash/before_algo.py
new file mode 100644
index 00000000..9baf6086
--- /dev/null
+++ b/codeflash/before_algo.py
@@ -0,0 +1,7 @@
+def postprocess(self, predictions: tuple[np.ndarray, ...], max_detections: int):
+    bboxes, logits = predictions
+    batch_size, num_queries, num_classes = logits.shape
+    logits_sigmoid = self.sigmoid_stable(logits)
+    for batch_idx in range(batch_size):
+        logits_flat = logits_sigmoid[batch_idx].reshape(-1)
+        sorted_indices = np.argsort(-logits_flat)[:max_detections]
diff --git a/codeflash/before_data.py b/codeflash/before_data.py
new file mode 100644
index 00000000..8134136a
--- /dev/null
+++ b/codeflash/before_data.py
@@ -0,0 +1,4 @@
+def check_missing_data(data: pd.DataFrame):
+    """Check if there is any missing data in the DataFrame"""
+    missing_data = data.isnull().sum().sum() > 0
+    return missing_data
diff --git a/codeflash/before_db.py b/codeflash/before_db.py
new file mode 100644
index 00000000..68ef3449
--- /dev/null
+++ b/codeflash/before_db.py
@@ -0,0 +1,6 @@
+def get_authors(session):
+    books = session.query(Book).all()
+    _authors = []
+    for book in books:
+        _authors.append(book.author)
+    return sorted(list(set(_authors)), key=lambda x: x.id)
diff --git a/codeflash/before_ml.py b/codeflash/before_ml.py
new file mode 100644
index 00000000..6449a3a5
--- /dev/null
+++ b/codeflash/before_ml.py
@@ -0,0 +1,11 @@
+import jax.numpy as jnp
+
+
+def sum_hand(hand):
+    """Returns the total points in a hand."""
+    return sum(hand) + (10 * usable_ace(hand))
+
+
+def usable_ace(hand):
+    """Checks to se if a hand has a usable ace."""
+    return jnp.logical_and((jnp.count_nonzero(hand == 1) > 0), (sum(hand) + 10 <= 21))
diff --git a/codeflash/before_numerical.py b/codeflash/before_numerical.py
new file mode 100644
index 00000000..ba6a19db
--- /dev/null
+++ b/codeflash/before_numerical.py
@@ -0,0 +1,22 @@
+def _equalize_cv(img: np.ndarray, mask: np.ndarray | None = None) -> np.ndarray:
+    histogram = cv2.calcHist([img], [0], mask, [256], (0, 256)).ravel()
+    i = 0
+    for val in histogram:
+        if val > 0:
+            break
+        i += 1
+    i = min(i, 255)
+
+    total = np.sum(histogram)
+    if histogram[i] == total:
+        return np.full_like(img, i)
+
+    scale = 255.0 / (total - histogram[i])
+    _sum = 0
+
+    lut = np.zeros(256, dtype=np.uint8)
+
+    for idx in range(i + 1, len(histogram)):
+        _sum += histogram[idx]
+        lut[idx] = clip(round(_sum * scale), np.uint8)
+    return sz_lut(img, lut, inplace=True)
diff --git a/codeflash/before_web_server.py b/codeflash/before_web_server.py
new file mode 100644
index 00000000..551fa58e
--- /dev/null
+++ b/codeflash/before_web_server.py
@@ -0,0 +1,12 @@
+async def get_endpoint(session, url):
+    async with session.get(url) as response:
+        return await response.text()
+
+
+async def some_api_call(urls):
+    async with aiohttp.ClientSession() as session:
+        results = []
+        for url in urls:
+            result = await get_endpoint(session, url)
+            results.append(result)
+        return results