diff --git a/requirements.txt b/requirements.txt
index 63a861e..d476d68 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -3,3 +3,4 @@ tqdm
 docker
 beautifulsoup4
 unidiff
+PyGithub
diff --git a/stats_pull_requests.py b/stats_pull_requests.py
index aaaed11..7182375 100644
--- a/stats_pull_requests.py
+++ b/stats_pull_requests.py
@@ -1,85 +1,68 @@
-import os, requests
+import os
 from datetime import datetime
 import pandas as pd
 import tqdm
+from github import Github
 
-COMMON_HEADERS = {
-    'Accept': 'application/vnd.github+json',
-    'Authorization': f'Bearer {os.environ["GITHUB_AUTH_TOKEN_CRAB"]}',
-    'X-Github-Api-Version': '2022-11-28',
-}
+# Initialize GitHub API client
+g = Github(os.environ["GITHUB_AUTH_TOKEN_CRAB"])
 
-def github_call(url, params = {}):
-    result = requests.get(url, headers=COMMON_HEADERS, params=params)
-    if result.status_code != 200:
-        raise Exception(f"Failed to fetch {url}: {result.status_code}, {result = }")
-    return result
-
-def get_pulls(repo_url: str) -> list[dict]:
-    response = github_call(f'{repo_url}/pulls', params={"state": "all"})
-    return response.json()
-
-def has_only_1_round_of_comments(commits: list[dict], comments: list[dict]) -> bool:
-    if len(comments) == 0 or len(commits) == 0:
+def has_only_1_round_of_comments(commits, comments):
+    if not comments or not commits:
         return False
-
-    # Convert timestamps to datetime objects for easy comparison
-    commit_dates = [datetime.fromisoformat(c["commit"]["author"]["date"]) for c in commits]
-    comment_dates = [datetime.fromisoformat(c["created_at"]) for c in comments]
+    
+    commit_dates = [c.commit.author.date for c in commits]
+    comment_dates = [c.created_at for c in comments]
+    
     commit_dates.sort()
     comment_dates.sort()
-
-    # Identify the first and last comment times
+    
     first_comment_time = comment_dates[0]
     last_comment_time = comment_dates[-1]
-
+    
     for commit_time in commit_dates:
         if first_comment_time < commit_time < last_comment_time:
             return False
-
+    
     return True
 
-
-def process_pull(repo_name: str, pull_number: str) -> dict:
-    pull = github_call(f"https://api.github.com/repos/{repo_name}/pulls/{pull_number}").json()
-    commits = github_call(f"https://api.github.com/repos/{repo_name}/pulls/{pull_number}/commits").json()
-    comments = github_call(f"https://api.github.com/repos/{repo_name}/pulls/{pull_number}/comments").json()
-
+def process_pull(repo, pull):
+    commits = list(pull.get_commits())
+    comments = list(pull.get_review_comments())
+    
     return {
-        "repo": repo_name,
-        "pr_number": pull["number"],
-        "additions": pull["additions"],
-        "deletions": pull["deletions"],
-        "changed_files": pull["changed_files"],
+        "repo": repo.full_name,
+        "pr_number": pull.number,
+        "additions": pull.additions,
+        "deletions": pull.deletions,
+        "changed_files": pull.changed_files,
         "has_only_1_round_of_comments": has_only_1_round_of_comments(commits, comments),
         "has_only_1_comment": len(comments) == 1,
     }
 
-def process_repo(repo_name: str) -> list[dict]:
+def process_repo(repo_name):
+    repo = g.get_repo(repo_name)
     stats = []
-    pulls = get_pulls(f"https://api.github.com/repos/{repo_name}")
-    for pull in tqdm.tqdm(pulls, desc=repo_name, leave=False):
-        if "merged_at" not in pull or pull["merged_at"] is None:
+    
+    for pull in tqdm.tqdm(list(repo.get_pulls(state="closed")), desc=repo_name, leave=False):
+        if not pull.merged_at:
             continue
-
-        stats.append(process_pull(repo_name, pull["number"]))
+        
+        stats.append(process_pull(repo, pull))
     return stats
 
-
 def main():
     repos = pd.read_csv("results.csv")
     repos = repos[repos["good_repo_for_crab"] == True]
-    print(len(repos))
     stats = []
-
-    for _, row in tqdm.tqdm(repos.iterrows(), total=len(repos)):
-        if "name" not in row or not isinstance(row["name"], str):
-            continue
-        name = row["name"]
-        stats.extend(process_repo(name))
-
-    pd.DataFrame(stats).to_csv("pr_stats.csv", index=False)
-
+    
+    try:
+        for _, row in tqdm.tqdm(repos.iterrows(), total=len(repos)):
+            if "name" not in row or not isinstance(row["name"], str):
+                continue
+            stats.extend(process_repo(row["name"]))
+    finally:
+        pd.DataFrame(stats).to_csv("pr_stats.csv", index=False)
 
 if __name__ == "__main__":
     main()