diff --git a/proposals/3014-full-event-with-rooms-pusher.md b/proposals/3014-full-event-with-rooms-pusher.md
new file mode 100644
index 00000000000..0f8f08961e9
--- /dev/null
+++ b/proposals/3014-full-event-with-rooms-pusher.md
@@ -0,0 +1,104 @@
+# MSC3014: HTTP Pushers for the full event with extra rooms information
+
+With [MSC2782](https://github.com/matrix-org/matrix-doc/pull/2782) adding a pusher for the full event
+content and [MSC3013](https://github.com/matrix-org/matrix-doc/pull/3013) allowing to encrypt push
+content huge steps are taken towards not needing to do any http requests at all to be able to process
+incoming push frames properly. The notable exception here are badge-only frames, as they only tell
+that the badge count updated, and its new count, but there is currently no way to tell which rooms
+specifically have been read to be able to dispose of the notification, apart from calling `/sync` or
+`/notifications`.
+
+## Proposal
+
+For pushers of type `http` (and thus all pushers inheriting from that type) a new `format`,
+`full_event_with_rooms`, is introduced. This pusher has the same fields as the `full_event` pusher
+described in [MSC2782](https://github.com/matrix-org/matrix-doc/pull/2782), with the addition of a
+new `rooms` field inside of the `counts` object. This new field is a map of room IDs to notification
+count. The rooms with a notification count of zero are not to be included in this map.
+
+As such, an example of an event being pushed out could look as following:
+
+```json
+{
+  "notification": {
+    "event_id": "$3957tyerfgewrf384",
+    "room_id": "!slw48wfj34rtnrf:example.com",
+    "type": "m.room.message",
+    "sender": "@exampleuser:matrix.org",
+    "sender_display_name": "Major Tom",
+    "room_name": "Mission Control",
+    "room_alias": "#exampleroom:matrix.org",
+    "prio": "high",
+    "content": {
+      "msgtype": "m.text",
+      "body": "I'm floating in a most peculiar way."
+    },
+    "counts": {
+      "unread": 4,
+      "missed_calls": 1,
+      "rooms": {
+        "!someroom:example.org": 3,
+        "!slw48wfj34rtnrf:example.com": 1,
+      }
+    },
+    "devices": [
+      {
+        "app_id": "org.matrix.matrixConsole.ios",
+        "pushkey": "V2h5IG9uIGVhcnRoIGRpZCB5b3UgZGVjb2RlIHRoaXM/",
+        "pushkey_ts": 12345678,
+        "data": {},
+        "tweaks": {
+          "sound": "bing"
+        }
+      }
+    ]
+  }
+}
+```
+
+And a badge count only update could look as following:
+
+```json
+{
+  "notification": {
+    "counts": {
+      "unread": 4,
+      "missed_calls": 1,
+      "rooms": {
+        "!someroom:example.org": 3,
+        "!slw48wfj34rtnrf:example.com": 1,
+      }
+    },
+    "devices": [
+      {
+        "app_id": "org.matrix.matrixConsole.ios",
+        "pushkey": "V2h5IG9uIGVhcnRoIGRpZCB5b3UgZGVjb2RlIHRoaXM/",
+        "pushkey_ts": 12345678,
+        "data": {},
+        "tweaks": {
+          "sound": "bing"
+        }
+      }
+    ]
+  }
+}
+```
+
+## Alternatives
+
+Instead of introducing a new pusher format a new flag to toggle the `rooms` field in the data of setting
+an http pusher could be added. That way, the new `rooms` field could also be used in the `event_id_only`
+format. As `event_id_only` needs to query `/sync`, `/event` or `/notifications` anyways it would not
+benifit from this MSC at all.
+
+## Security considerations
+
+If not used in combination of [MSC3013](https://github.com/matrix-org/matrix-doc/pull/3013) this will
+leak even more data to an untrusted gateway.
+
+As one might create their own custom trusted gateway with e.g. gotify that could benefit from this,
+this pusher format is also available for the plaintext http pusher.
+
+## Unstable prefix
+
+Is this needed for pusher formats?