Hook panel A/B tester
A shared demo that scores TikTok and Instagram hooks against about 100 personas in near real time with Jev.
This summary is based on how curated lists describe the project; its README was not available to review.
Overview
This entry is a post on X showing a hook A/B testing panel for short-form video creators. A draft hook for TikTok or Instagram is scored against roughly 100 audience personas in near real time, giving a quick read on which personas the opening would grab.
The idea relies on decision-model calls being fast and cheap enough to fan out across many personas per keystroke or draft. No repository or README was available, so the scoring rubric, persona definitions and aggregation method are not documented.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Score | How strongly would this persona engage with this hook?(inferred) | — |
How it works
- Creator writes a draft hook
- Each of about 100 personas is paired with the hook as a Jev question
- Scores are aggregated into a panel view for comparing variants
Strengths
- Fan-out across many personas is feasible because each call is small and fast
- Gives a structured comparison between hook variants
Caveats
- Only a social post; no code available
- Persona scores are model judgments, not real audience data
At a glance
- Type
- Discussion
- Status
- discussion
Use cases and integrations
How curated lists describe it
Near-real-time scoring of TikTok and Instagram hooks against about 100 personas.
hellogumbo/awesome-jev
Frequently asked questions
What does the panel decide?
How each of about 100 personas would respond to a video hook, as a score.
Is the code available?
No public repository was linked; the entry is an X post.
Can persona scores replace real A/B tests?
No; they are a fast pre-screen of drafts, not measured audience behaviour.
Related projects
Cloudflare app that labels a YouTube creator's Shorts with Laya or Jev to find which hooks drive views and score new hooks.
ProjectprototypeTypeScriptSelf-hostable Reddit buyer-intent finder that has Jev judge every post and comment it reads, with a written reason and cost per lead.
ProjectproductionTypeScript- jev-seo★ 610
Live SEO audit from one homepage URL: crawl, 52 rules, Core Web Vitals and Jev page judgments, reported as PDF, XLSX and Markdown.
PluginbetaPython Scores HeyReach LinkedIn leads and first-message variants with Jev, picks a message per lead and writes draft campaigns back.
ProjectbetaPythonSite audit for SEO and AI-search visibility: a deterministic crawl, narrow Jev Choice, Score and Noul questions, and a live dashboard.
AppbetaTypeScriptCLI that has Jev judge a video's subtitles segment by segment, then cuts a highlight reel and a filler-free full version.
ProjectbetaPython