WEBVTT

00:00:00.341 --> 00:00:03.499
This walkthrough follows a real Ruthless Editor analysis of the

00:00:03.541 --> 00:00:05.931
opening of Death Takes a Honeymoon by Kat Summers

00:00:06.571 --> 00:00:09.429
The app captures are edited into steps with narration added

00:00:09.557 --> 00:00:09.941
afterward

00:00:10.425 --> 00:00:11.620
This is a short excerpt review

00:00:12.217 --> 00:00:14.137
not a claim that the whole book has been edited

00:00:14.820 --> 00:00:17.678
Begin with one passage and give the saved manuscript a

00:00:17.721 --> 00:00:18.233
useful name

00:00:18.788 --> 00:00:20.196
Keep your original file separately

00:00:20.813 --> 00:00:22.520
Browser saves belong to this browser

00:00:22.989 --> 00:00:25.037
so export a backup if you want to move devices

00:00:25.464 --> 00:00:27.640
or protect your work before clearing browser data

00:00:28.280 --> 00:00:30.371
Choose your writing provider and an available model

00:00:30.883 --> 00:00:32.333
then enable Use saved connection

00:00:33.155 --> 00:00:35.715
This run used OpenAI with the model shown on screen

00:00:36.142 --> 00:00:37.379
Your key stays out of the demo

00:00:37.891 --> 00:00:41.347
Analysis uses your provider account and a longer manuscript can

00:00:41.389 --> 00:00:42.584
require more than one request

00:00:43.181 --> 00:00:45.272
The editorial brief tells the model what matters

00:00:46.012 --> 00:00:46.226
Here

00:00:46.695 --> 00:00:48.914
the genre is supernatural romantic comedy

00:00:49.468 --> 00:00:52.370
The instructions preserve Corvus's comic first person voice

00:00:52.882 --> 00:00:53.820
deliberate fragments

00:00:54.290 --> 00:00:54.887
exaggeration

00:00:55.442 --> 00:00:56.380
and the repeated phrase

00:00:56.850 --> 00:00:57.362
Four days

00:00:58.107 --> 00:01:00.027
Protected phrases add a firmer boundary

00:01:00.539 --> 00:01:04.720
proposals overlapping those phrases cannot be accepted Run the analysis

00:01:05.104 --> 00:01:05.830
then check coverage

00:01:06.384 --> 00:01:09.286
This passage contains ninety nine words in one segment

00:01:09.866 --> 00:01:12.468
The completed status means that segment was analyzed

00:01:12.980 --> 00:01:14.986
It does not mean every sentence needs changing

00:01:15.498 --> 00:01:17.290
or that every model suggestion is right

00:01:17.930 --> 00:01:20.874
The actual response offered one punctuation suggestion

00:01:21.501 --> 00:01:23.080
It replaced the parenthetical aside

00:01:23.549 --> 00:01:24.061
lovely bones

00:01:24.573 --> 00:01:26.152
that one with em dashes

00:01:26.749 --> 00:01:29.352
Use View source passage to read that line in context

00:01:29.949 --> 00:01:30.589
Compare the original

00:01:31.144 --> 00:01:33.576
replacement and explanation before deciding

00:01:34.141 --> 00:01:36.232
Accepting the proposal updates the edited draft

00:01:36.744 --> 00:01:38.408
It does not overwrite the imported source

00:01:39.048 --> 00:01:42.291
Undo decision restores the original wording and returns the proposal

00:01:42.334 --> 00:01:42.632
to pending

00:01:43.230 --> 00:01:45.363
In this review I then reject the change

00:01:46.148 --> 00:01:48.921
The parenthetical pause already suits the narrator's voice

00:01:49.561 --> 00:01:51.396
I add an author note explaining that choice

00:01:52.078 --> 00:01:55.364
That is a valid editorial outcome A suggestion can be

00:01:55.406 --> 00:01:57.966
reasonable and still be unnecessary for this passage

00:01:58.540 --> 00:02:02.124
Preserve purposeful style rather than accepting changes simply to make

00:02:02.167 --> 00:02:03.319
the proposal count go down

00:02:03.874 --> 00:02:04.898
Export the review report

00:02:05.410 --> 00:02:06.050
edited draft

00:02:06.519 --> 00:02:07.117
original source

00:02:07.586 --> 00:02:08.653
and editable backup

00:02:09.122 --> 00:02:10.615
These files serve different purposes

00:02:11.313 --> 00:02:12.550
The report records decisions

00:02:13.105 --> 00:02:14.555
the backup restores the workspace

00:02:15.110 --> 00:02:17.670
the text exports let you compare the result outside the

00:02:17.713 --> 00:02:17.841
app

00:02:18.310 --> 00:02:21.212
Original and edited text matched in this run because the

00:02:21.254 --> 00:02:22.534
only proposal was rejected

00:02:23.087 --> 00:02:25.860
Read the final export before replacing anything in your writing

00:02:25.988 --> 00:02:26.159
project

00:02:26.884 --> 00:02:28.932
Keep the current manuscript and your review history

00:02:29.772 --> 00:02:31.863
The goal is a deliberate decision you can explain

00:02:32.545 --> 00:02:34.892
with your source and your voice still under your control