Edge spent $0 on marketing. usage more than doubled last week.
you ask AI to price your product. you get $9, $29, $99. an expert asks what your customers pay today to solve this.
Edge makes every AI do a task the way the best person in the world does it, and that person gets paid each time.
monday alone: 584 uses, our biggest day so far.
Elena Verna leads growth at Lovable, which hit $200M ARR in under a year.
so we turned her public writing into a Claude growth prompt.
not 50 growth hacks. one loop, the missing numbers, one next test.
comment LOOPS for the copyable prompt. first 20 get 30 days of Edge Pro.
haiku 5.5 + skills beat opus 5.5 without them.
20 founder tasks, 240 blind AI verdicts:
haiku + skill: 88 wins
opus: 75
ties: 77
haiku led on 10 tasks. opus on 7.
the cheapest claude, with the right expertise, beat opus without it.
https://t.co/0wZandRicF
Opus 5.5 built this video in Claude Code. One skill, no video editor.
Hook on frame 0, every cut on the beat, its own score from code.
The skill is free, link below.
we turned the workflow into a skill so anyone can reproduce it.
Edge automatically gives your AI the right skills for each task.
try this exact skill:
https://t.co/ZOEaoeOXUV
we needed 4 Edge setup videos. an ai agent recorded all 4 on a real Mac.
it drove each setup, recorded the run, and turned its own action log into captions.
zero human clicks.
but sure, keep recording Looms.
i'm 25 and i just cloned myself.
it knows the monastery story, the hinge disaster, and the book i make everyone read.
only 100 people get a copy.
i might regret this. but i do this for humanity.
https://t.co/plswIQhpvo
same claude. ~50% more tasks solved.
edge searches 130k+ public skills for the right expert playbook and plugs it into your ai automatically. 110 ms.
20 pro keys ($99/mo) are reserved for replies to this post.
reply "edge" and we'll dm you one.
we gave claude 1 slide and 130,000 expert playbooks.
we didn't tell it which one to use.
it picked one in 110 ms. the slide went from "redo this" to the board deck.
same model. same prompt. only the playbook changed.
https://t.co/XO5KdTBhTS
Introducing Edge 🚀
Curated skills make agents solve ~50% more tasks.
Spotify-style ranking for your agent's skills.
130k+ indexed. Daily trend scrapes. Jev-powered ranking. Learns from user ratings. Failed security scans get withheld.
Claude Code, Codex, Cursor.
https://t.co/EORysFUwXy
ChatGPT with no skill beat Remotion.
We tested 10 video skills.
No skill: 55
Remotion Official Skills: 31
The winner: 90
Popular ≠ best.
Edge tests the skills before your agent uses them.