Juan B.

@andi_joh 💖 112218

[Juan B. @j_spam]

@fyreinteractive scripts

[Juan B. @j_spam]

@Polymarket based

[Juan B. @j_spam]

@fyreinteractive Ohh i love it
I definitely see the difference in scripts between normal claude opus and using facelessos
And just in general way better scripts to begin with

[Juan B. @j_spam]

@fyreinteractive Gemini

[Juan B. @j_spam]

@fyreinteractive Ai

[Haris @fyreinteractive]

opus 4.8 dropped and you can now run an ENTIRE faceless youtube channel from ONE chat...

just connect Nexlev + ElevenLabs + Canva + inVideo AI into claude with MCPs

and voilá:

> one prompt does the research
> voiceover, thumbnail and edit, no tab-switching
> the channel even self-corrects after it learns from its own analytics

and i've decided to document the ENTIRE autonomous stack...

the exact 6-tool MCP setup that takes a niche from research to a published video without you ever leaving the claude window

the same way channels printing $10k-$100k+/month run their whole operation, fully automated

here's what's included inside the doc:

- the NexLev research MCP

(finds faceless channels under 50k subs suddenly pulling outlier views, rips their audience-retention drop-off curves, and estimates RPM per niche so you build in a $15 RPM niche instead of a $3 one)

- claude opus 4.8 as the script brain

(reverse-engineers the packaging FIRST, 10 high-CTR title + thumbnail concepts before a word of script exists, then writes around the EXACT retention drop-offs the data exposed and front-loads a pattern interrupt where competitors bleed 30% in the first 20 seconds)

- the elevenlabs voice MCP

(cloned voice with dynamic vocal fry, real breathing and pacing mapped to the emotional arc, tense sections faster and revelations slower, broadcast-quality audio dropped straight back into the chat)

- the canva thumbnail MCP

(autofills the thumbnail from your locked brand template using the title text generated back in research, exports the high-res png into the conversation while you do nothing)

- invideo + agent one for the render

(picks the best video model per scene, photorealistic footage, AI cinematic, or stock b-roll, swaps elements on command, syncs to the elevenlabs audio and renders the final mp4)

- the youtube api for publish + self-correction

(auto-publishes with SEO description and tags scheduled for peak hours, then 48 hours later pulls CTR, retention and RPM back so claude permanently adjusts its thumbnail and scripting logic for the next video)

all backed by 7,000+ faceless youtube videos engineered across 50+ niches that lead to 150M+ views

and for 24h, it can be all yours for free

like + comment "AI" and i'll send it over

(must be following + RT for priority access)

[Juan B. @j_spam]

@fyreinteractive Ai

[Juan B. @j_spam]

@RealCynicalFox Imagine going to fight and die for Israel and then getting paid shit too

[Juan B. @j_spam]

@InfernoYTA i can just do this on capcut for cheaper, whats the incentive here

[Juan B. @j_spam]

@iamwilsxn tiktok does not pay more what crack are you smoking

[Juan B. @j_spam]

@primalrobin 11000

[Juan B. @j_spam]

@mister_manners_ Ai ahh tweet

[Juan B. @j_spam]

@bokoen1 i love you bo

[Juan B. @j_spam]

@1of10media This video was so shitty

[Juan B. @j_spam]

@NashJohn22 Hey bro im also doing history videos
Would live to chat with you

[Juan B. @j_spam]

@noahmorris Let me back into vidrush i was to broke to afford it when i got my invite 😭

[Juan B. @j_spam]

@NashJohn22 Hey bro what was your prompt for the image if you dont mind sharing

[Juan B. @j_spam]

@Richard_YTS Two of those are mine lol

[Juan B. @j_spam]

@TWarPhilosopher @cynomel A fight?