diff --git a/README.md b/README.md
index 97b0530..912c366 100644
--- a/README.md
+++ b/README.md
@@ -4,13 +4,20 @@ This is a toy I built so I wouldn't have to check the [Grand Forks 311](http://w
 app/website all the time.
 
 It currently toots to Mastodon at [@hackgfk_311@botsin.space](https://botsin.space/@hackgfk_311).
-Toots are converted to Tweets ([@hackgfk_311](https://twitter.com/hackgfk_311)) with @renatolond's wonderful 
+Toots are converted to tweets ([@hackgfk_311](https://twitter.com/hackgfk_311)) with @renatolond's wonderful 
 crossposter: https://crossposter.masto.donte.com.br ([source code](https://github.com/renatolond/mastodon-twitter-poster)).
 
 Written in R because I'm much quicker at that than Python, and it seemed like a fun use case.
 
 Pull requests always welcome!
 
+# 311_ebooks
+
+In Python (because why stay consistent?), this script takes the corpus out of the local 311 database
+and builds an [ebooks-style](https://en.wikipedia.org/wiki/Horse_ebooks) toot using Markovify.
+
+Currently tooting at [@hackgfk_311_ebooks@botsin.space](https://botsin.space/@hackgfk_311_ebooks).
+
 -----
 # Hack Grand Forks 
 
diff --git a/auth.ini.default b/auth.ini.default
new file mode 100644
index 0000000..b09e50b
--- /dev/null
+++ b/auth.ini.default
@@ -0,0 +1,5 @@
+[mastodon]
+#needs trailing slash
+server = https://botsin.space/
+email = heck@yeah.com
+password = ReallyGoodPassword
diff --git a/bin/311_ebooks.py b/bin/311_ebooks.py
new file mode 100644
index 0000000..23d2f29
--- /dev/null
+++ b/bin/311_ebooks.py
@@ -0,0 +1,61 @@
+
+
+import markovify #https://github.com/jsvine/markovify
+import sqlite3 #to read db
+from mastodon import Mastodon
+from configparser import ConfigParser
+import os.path
+
+# Open the db
+conn = sqlite3.connect('requests.sqlite')
+c = conn.cursor()
+
+def get_descriptions():
+    c.execute('SELECT description FROM requests')
+    data = c.fetchall()
+    return(data)
+
+# Get descriptions and convert to strings from tuples
+b = ["".join(str(x).replace("\\r\\n","")) for x in get_descriptions()]
+
+# Build the model.
+text_model = markovify.Text(b)
+
+# Print three randomly-generated sentences of no more than 280 characters
+#for i in range(3):
+#    print(text_model.make_short_sentence(280))
+
+# Next up use this to toot: https://github.com/halcy/Mastodon.py
+
+# parse existing file
+config = ConfigParser()
+config.read('auth.ini')
+
+# read values from a section
+server = config.get('hackgfk_311_ebooks', 'server')
+email = config.get('hackgfk_311_ebooks', 'email')
+password = config.get('hackgfk_311_ebooks', 'password')
+
+# Register the app (once)
+if not os.path.isfile('hackgfk_311_ebooks_clientcred.secret'):
+    Mastodon.create_app(
+         'hackgfk_311_ebooks',
+         api_base_url = server,
+         to_file = 'hackgfk_311_ebooks_clientcred.secret'
+    )
+
+# Log in
+mastodon = Mastodon(
+    client_id = 'hackgfk_311_ebooks_clientcred.secret',
+    api_base_url = server,
+)
+mastodon.log_in(
+    email,
+    password,
+    to_file = 'hackgfk_311_ebooks_usercred.secret'
+)
+
+# Send toot
+toot = text_model.make_short_sentence(280)
+print(toot)
+mastodon.toot(toot)