diff --git a/requirements.txt b/requirements.txt
index 6b62a17..1a151bd 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,2 +1,3 @@
+python-dateutil==2.8.1
 todoist-python==8.1.1
 urlextract==0.14.0
diff --git a/todoist-github/cli.py b/todoist-github/cli.py
index 767f68b..b3bba8f 100755
--- a/todoist-github/cli.py
+++ b/todoist-github/cli.py
@@ -1,21 +1,71 @@
 #!/usr/bin/env python3
-from .clients import todoist
-from .utils import get_github_task
+import datetime
+
+from dateutil.relativedelta import relativedelta
+
+from .clients import github, todoist
+from .utils import get_github_issue_details, get_github_task, get_issue
+
+
+def get_issue_link(issue_or_pr) -> str:
+    return "[#{id}]({url})".format(id=issue_or_pr.number, url=issue_or_pr.html_url)
+
+
+def issue_to_task_name(issue) -> str:
+    return get_issue_link(issue) + ": " + issue.title
 
 
 def get_relevant_todoist_tasks():
     todoist.items.sync()
     tasks = {}
     for task in todoist.items.all():
-        if get_github_task(task["content"]):
-            tasks[task['content']] = task
+        github_task = get_github_task(task["content"])
+        if github_task:
+            tasks[github_task] = task
     return tasks
 
 
+def is_task_completed(task):
+    return task.data.get("checked", 0)
+
+
 def main():
     todoist_tasks = get_relevant_todoist_tasks()
+    relevant_since = datetime.datetime.now() - relativedelta(
+        weeks=30
+    )  # TODO: Make this a sane number
+    tasks_actioned = []
+    me = github.get_user()
+    for assigned_issue in me.get_issues(state="all", since=relevant_since):
+        task = todoist_tasks.get(assigned_issue.html_url)
+        if not task and assigned_issue.state == "open":
+            task = todoist.items.add(issue_to_task_name(assigned_issue))
+        if not task:
+            continue
+        tasks_actioned.append(task["id"])
+        if assigned_issue == "closed" and not is_task_completed(task):
+            print("completing", assigned_issue)
+            task.complete()
+        if is_task_completed(task):
+            print("uncompleting task", assigned_issue)
+            task.uncomplete()
+        if task["content"] != issue_to_task_name(assigned_issue):
+            print("updating issue name for", assigned_issue)
+            task.update(content=issue_to_task_name(assigned_issue))
+        if assigned_issue.milestone and assigned_issue.milestone.due_on:
+            task.update(
+                date_string=assigned_issue.milestone.due_on.strftime("%d/%m/%Y")
+            )
 
-
+    for task in todoist_tasks.values():
+        if not is_task_completed(task) or task["id"] in tasks_actioned:
+            continue
+        org, repo, issue_number = get_github_issue_details(task["content"])
+        issue = get_issue(me, org, repo, issue_number)
+        me_assigned = me.login in {assignee.login for assignee in issue.assignees}
+        if not me_assigned:
+            print("Deleting", issue)
+            task.delete()
 
 
 if __name__ == "__main__":
diff --git a/todoist-github/clients.py b/todoist-github/clients.py
index 5a31eec..e06f69e 100644
--- a/todoist-github/clients.py
+++ b/todoist-github/clients.py
@@ -1,4 +1,7 @@
-from todoist import TodoistAPI
 import os
 
+from github import Github
+from todoist import TodoistAPI
+
 todoist = TodoistAPI(os.environ["TODOIST_TOKEN"])
+github = Github(os.environ["GITHUB_TOKEN"])
diff --git a/todoist-github/utils.py b/todoist-github/utils.py
index 6eb62ce..27aa207 100644
--- a/todoist-github/utils.py
+++ b/todoist-github/utils.py
@@ -1,7 +1,11 @@
+import re
 from typing import Optional
-from urlextract import URLExtract
 from urllib.parse import urlparse
 
+from github.Issue import Issue
+from urlextract import URLExtract
+
+GITHUB_ISSUE_PR_RE = re.compile(r"\/(.+?)\/(.+?)\/(pull|issues)\/(\d+?)$")
 
 extractor = URLExtract()
 
@@ -10,5 +14,24 @@ def get_github_task(content) -> Optional[str]:
     if "github" not in content.lower():
         return None
     for url in extractor.gen_urls(content):
-        if urlparse(url).netloc == "github.com":
+        parsed_url = urlparse(url)
+        if parsed_url.netloc == "github.com" and GITHUB_ISSUE_PR_RE.search(
+            parsed_url.path
+        ):
             return url
+
+
+def get_github_issue_details(content):
+    url = get_github_task(content)
+    if not url:
+        return
+    parsed_url = urlparse(url)
+    match = GITHUB_ISSUE_PR_RE.search(parsed_url.path)
+    return match.group(1), match.group(2), match.group(4)
+
+
+def get_issue(me, org, repo, issue_num):
+    headers, data = me._requester.requestJsonAndCheck(
+        "GET", f"/repos/{org}/{repo}/issues/{issue_num}"
+    )
+    return Issue(me._requester, headers, data, completed=True)