ShortsDoctor: fix the first 3 seconds and the script behind it

Paste a script (or just an idea) for a TikTok, Instagram Reel or YouTube Short. ShortsDoctor tells you what is weak, shows the evidence from your own lines, and hands back a better version you can film today.

Fictional sample script. Spoken timing is estimated by the bundled analyzer; it is not platform analytics.

What you get in every run

• Scorecard (0-100 rubric score) across six dimensions: promise clarity, tension, specificity, speed to value, pacing, ending. Every score quotes a line from your script.
• Measured facts from a bundled script: estimated spoken length, seconds until your first line ends, greeting-style openers, filler density, early "follow/subscribe" asks, long flat stretches, repeated phrases.
• Top 3 fixes, ordered, each tied to a line.
• Five hook rewrites using five different patterns, each with word count, estimated seconds, and what your video must show to pay it off.
• Tightened script with timing, on-screen text and a visual cue per line.
• Caption file (.srt) for the rewritten script, in English, Korean or Japanese.
• Before/after check by re-running the same measurements on the rewrite.
• A/B test plan and a title, hashtags and pinned comment.

Send any of these

1. A finished script or voice-over text
2. A pasted transcript or caption export of a video you already posted
3. One line of idea ("30-second video about cold brew mistakes")

Good to know (honest limits)

• Works from text. It cannot watch a video or open a link. Paste the spoken text.
• It does not predict views, retention percentages or reach. Scores rate the craft of the script. Real results depend on your audience, topic, posting and the platform.
• Timing is estimated from word counts and a speaking speed (you can tell it your pace).
• It will not write hooks that promise something your video does not show, and it removes guarantee-style claims about health, money and results.
• One script per request works best. Follow-up edits are welcome.

Languages: scripts and replies in English, Korean and Japanese are measured with language-specific rules. Other languages are rewritten, but timing is approximate.