diff --git a/README.md b/README.md
new file mode 100644
index 0000000..8203d7a
--- /dev/null
+++ b/README.md
@@ -0,0 +1,38 @@
+# Custom Filter Framework
+
+## Rules
+A single rule should be able to match multiple things and have multipl
+actions, i.e.:
+
+### Example
+```
+Hdr1: regexp
+Hdr2: regexp
+<action1>
+<action2>
+```
+
+### Misc
+ * Pseudo header for To/CC.  Some way to say "directly to me" that works with
+BCC.
+
+## Use cases
+ * Simple `To: me@example.com` -> add label `@Me`
+ * `From: wife` only `To: me` vs `From: wife` & `To: multiple people`
+ * `To: important CC: unimportant` still gets `@Important`
+ * Mix AND, OR and negation in a single rule.
+
+## Actions
+ * Add label
+ * Mark read
+
+# Threading
+ jwz or something simple first?  gmail has a nice property of being able to
+ see message / thread ID by default, and then fetch by message or thread ID.
+ Fetch by thread ID only way to do batch processing via API.
+
+# Old Notes
+ * Add `fast_header` table that is just the values we`d want to search on
+ (i.e.  the headers we hash on, plus whatever is needed for jwz threading).
+ * Investigate stored procedures that update `fast_header` automatically when
+ we we add bulk headers to table `header`
diff --git a/TODO b/TODO
deleted file mode 100644
index d248dd1..0000000
--- a/TODO
+++ /dev/null
@@ -1,4 +0,0 @@
-- Add 'fast_header' table that is just the values we'd want to search on (i.e.
-    the headers we hash on, plus whatever is needed for jwz threading).
-- Investigate stored procedures that update 'fast_header' automatically when
-    we we add bulk headers to table 'header'