diff --git a/TODO.md b/TODO.md
new file mode 100644
index 0000000..ed65d3d
--- /dev/null
+++ b/TODO.md
@@ -0,0 +1,24 @@
+# TODO
+
+## Next Steps:
+
+1. Define the status of services (OK and ERROR enough?)
+2. Handle the status of services and produce a new internal status (OK, ERROR, WARNING enough?)
+3. Implement the Task interface and write a task to store output in a database
+4. Implement the Plugin interface and write a plugin to show some data and statistics in a web interface
+
+## Things to do in no particular order:
+
+- Implement error handling at all needed places! currently nearly no error handling is done!
+- Configure the logger
+- Make use of the logger wherever it us useful! and be informative when information really helps debugging the application!
+- Multicore support when scheduling jobs with APScheduler
+- Parent alerting (escalation)
+- Core error alerting to [linspector]->members on internal errors
+- Hostgroup parents (example: if the hostgroup "network" is down, don't alert for the hosts in that network)
+- Add more services
+- Add more tasks e.g. storage in mongodb or mariadb and some more notifications like XMPP and SMS
+- Check for all required configuration options and set defaults in services.
+- Write documentation and inline documentation.
+- Add kwargs to notifications, tasks and maybe plugins.
+- Add date and cron based scheduling to linspector.py to make the full use of APScheduler.