Fer Zenth

[Fer Zenth @steepbrooo]

@IPXExchange my journey using the IPX app so far:

First day I opened it, did the Fan Check, picked my fandom, and got my Fan Card. Felt like joining a club

Then I started posting. My first ones sounded like a press release ALFRED rejected them, three times. That stung, but it taught me the rule: sound like a fan, not an announcement.

So I dug into how $IPX actually works the router, the escrow, the way creators get paid and wrote it in my own words. That thread scored big.

Now it's a routine: check the leaderboard, post, keep the streak alive. 2.3x became 2.4x, now 2.7x. Still outside the top 100, but I'm closer than I was. Launch is near

The app didn't just give me a card. It gave me a reason to actually understand what I'm backing

[Fer Zenth @steepbrooo]

@IPXExchange I finally tried buying into a project through the Atomic Router, and I get it now.

One click, one transaction. It took the 0.5% fee, routed through the IPX/USDC pool on Aerodrome, split my money into the exact USDC the production still needed plus $IPX for market liquidity, and the project tokens landed in my wallet. Whatever was left over came straight back to me, same asset.

The part that sold me: if any single step fails, the whole thing reverts. Nothing half-done, no money stuck halfway.

And while the raise is live, my USDC isn't floating around unbacked it sits fully backed in the curve, refund or graduate. Even the idle USDC earns Coinbase-funded rewards that feed the Protocol Vault, so the capital isn't just sitting there doing nothing.

Capital protection plus automatic liquidity at graduation, in one click. That's the real product

[Fer Zenth @steepbrooo]

found @onlystreetsfun on Rarelist and the vibe checks out. catch a live moment with your phone, name it, tokenize it in seconds. no gallery uploads, no recycled pics just what you actually saw. top 200 creators get $USDT + OnlyStreets at TGE. your street could be next

[OnlyStreets @onlystreetsfun]

Introducing https://t.co/YJfAG0wpgd

A Solana launchpad for fresh authentic memes born in the real world.

Capture something iconic IRL, give it a name, and launch a coin.

Coins can ONLY be created from photos taken directly in the app.

[Fer Zenth @steepbrooo]

@onlystreetsfun wow this amazing

[Fer Zenth @steepbrooo]

@IPXExchange my FAN SZN journey, start to now:

I came in thinking a good post meant sounding official. ALFRED gave me nothing for it a tag it couldn't read, then three rejections for sounding like an announcement

so I sat down and learned how $IPX actually works. the long thread that came out of it scored big. that was the turn.

now my streak is 2.7x, I'm still outside the top 100, and I'm writing from my own experience instead of a feature list. launch is two days away. I'm not done yet

[Fer Zenth @steepbrooo]

@ZeroGuysNFT @LayerZero_Core let's join with me @ambassador_grim @BJauhi @manhhuynh2310

[Zero Guys @ZeroGuysNFT]

First NFT collection on @LayerZero_Core

Apply for WL : https://t.co/D9nNn0cYP3

Applications closing soon.

[Fer Zenth @steepbrooo]

@IPXExchange the quietest part of FAN SZN for me: the streak

it started at 2.3x. then 2.4x. now it's 2.7x. nothing I posted changed the math as much as simply showing up again the next day.

one great post at 90+ is worth up to ~2,700 AURA now, but only because the streak is there behind it. the multiplier is the reward for not quitting.

that's the part nobody screenshots: the boring days between the big posts

[Fer Zenth @steepbrooo]

@IPXExchange my FAN SZN experience so far: cheering doesn't pay here
homework does

the post that scored big for me wasn't hype it was the long one where I sat down and explained how $IPX actually works, piece by piece

then I tried shortcuts. three rejections in a row. lesson learned: if I can't explain it in my own words, I don't understand it yet.

I'm still outside the top 100, posting most days, streak building. that's the honest version of being a fan here not five posts a day for the sake of it, but actually getting what you're backing

[Fer Zenth @steepbrooo]

.@IPXExchange the hardest part of making a film? it's not always the filming. the Production Studio Suite is built for everything after the cut: rights, tax-credit paperwork, rebates, asset organization.

deep dive into their Production Studio Suite:
• built-in tax-credit and rebate engine
• on-chain rights ledger for each title
• asset vaults next to the filmmaker tools

rights records and project assets live with the production workflow instead of scattered across disconnected tools.

i think managing a project doesn't stop at the final cut. creators still need to track ownership and handle the business side.

my take is production tools should make everything after the shoot easier to manage. the creative work deserves a better system behind it

[Fer Zenth @steepbrooo]

@T9090z yes

[Fer Zenth @steepbrooo]

I just got onboard @VOICECC Express! Its your LAST call! Ends in 6 days!

[Fer Zenth @steepbrooo]

@IPXExchange the IPX part I keep coming back to: the on-chain settlement engine. because here's the thing nobody talks about making money from IP is only half the job. the other half is getting paid, on time, transparently. and that's where most creators get screwed.

think about how many ways a creator actually earns: streaming revenue, licensing, sync deals, ticket sales, merch. five different pipes, five different payout schedules, five different spreadsheets. try tracking where your money is coming from good luck.

the settlement engine is IPX's answer to that mess: it monitors all those revenue streams in one place, on-chain, and distributes rewards to the relevant creators in real time. no more waiting months for a payout cycle to close. no more reconciling spreadsheets line by line

and here's why it matters more than it sounds: it plugs into the rest of IPX. a deal gets created, a creator gets matched, a project gets produced and when the revenue comes back, the same engine routes it to the people who made it. one loop, end to end.

that's the part worth watching, honestly. everyone in crypto wants to talk about generating value. very few want to talk about distributing it fairly. a better creator economy needs better ways to pay people, not just make money. IPX is building that second half

[Fer Zenth @steepbrooo]

@IPXExchange the IPX part I keep coming back to: the on-chain settlement engine. because here's the thing nobody talks about making money from IP is only half the job. the other half is getting paid, on time, transparently. and that's where most creators get screwed.

think about how many ways a creator actually earns: streaming revenue, licensing, sync deals, ticket sales, merch. five different pipes, five different payout schedules, five different spreadsheets. try tracking where your money is coming from good luck.

the settlement engine is IPX's answer to that mess: it monitors all those revenue streams in one place, on-chain, and distributes rewards to the relevant creators in real time. no more waiting months for a payout cycle to close. no more reconciling spreadsheets line by line

and here's why it matters more than it sounds: it plugs into the rest of IPX. a deal gets created, a creator gets matched, a project gets produced and when the revenue comes back, the same engine routes it to the people who made it. one loop, end to end.

that's the part worth watching, honestly. everyone in crypto wants to talk about generating value. very few want to talk about distributing it fairly. a better creator economy needs better ways to pay people, not just make money. IPX is building that second half

[Fer Zenth @steepbrooo]

@IPXExchange the Bridge news is the IPX update that actually changed how I see the launch. real revenue sponsorships, ticket sales flowing on-chain as USDC from Nov 15. if that pipe works, the tokens follow real money, not just attention. that's the part I'm watching

[Fer Zenth @steepbrooo]

1/ @IPXExchange IPX vs the crowdfunding you already know Kickstarter, Indiegogo. same idea on the surface (fans fund a project), completely different machine underneath. here's the difference that matters.

2/ on Kickstarter you donate and wait. your money leaves the moment you pledge, the creator gets it, and if the project dies you get an apology email. no refund, no resale, no stake.

3/ on IPX the money goes into milestone escrow. it's released stage by stage pre-production, filming, post only after proof of work. raise fails? refunded pro-rata. project stalls? surrender tokens for fixed USDC.

4/ and after the raise, your backing is a token, not a receipt. it trades on the exchange through Atomic Router one transaction, 0.5% fee, leftover USDC back. you can exit, not just wait and hope.

5/ so it's not crowdfunding with a crypto coat of paint. it's funding where the money is held to account, and the backing itself has a market. that's the difference

[Fer Zenth @steepbrooo]

.@IPXExchange ok, I finally tried it myself instead of just reading about it. picked a fandom, did the Fan Check, got my Fan Card. whole thing took a few minutes

2/ what stuck with me: nothing about it feels like a black box. the raise, the escrow, the milestones you can see where the money sits and when it moves. that's the part I didn't expect to care about, but I do

[ryu 龙 @0xgoryu]

Alhamdulillah berkah rabu pagi ❤ lumayan dapat 15+ temuan bug low-high severity di program self hosted

moga aja dari sekian temuan ada yang nyangkut dan diacc

nih gw kasih prompt langsung tempel di google search dan kalian bisa research sendiri:

1. dork prioritas

("Bug Bounty Program" | ("Vulnerability" & "Reward")) -bugcrowd -hackerone -yeswehack -intigriti -immunefi

2. dork basis security.txt

inurl:/.well-known/security.txt "reward" ("$" OR "€" OR "BTC")

3. dork frasa spesifik

"responsible disclosure" ("cash reward" OR "monetary reward") -hackerone -bugcrowd

4. dork multi language

site:*.de inurl:/sicherheit intext:belohnung
site:*.fr ("prime de sécurité" OR "divulgation responsable")

filter waktu kalo mau spesifik:

klik Tools → Any time → pilih Past month (untuk 1–4 minggu) atau Past week (untuk 1 minggu terakhir)

thanks me later

[Bruno @BrunoF566]

Claude animations are F*CKING EXPLODING

every video is made in 5 minutes with a single prompt

and this MF racked up over 1B views with them

So I made my own and decided to leak the system

Just paste this into Claude and enjoy

<role>
You are a motion-graphics engineer and animator. You build wordless "contrast parable" videos entirely with code: two chunky white pictogram characters on a black, glossy, reflective stage. No AI video model, stock footage or image generator.
</role>

<task>
Turn the script/idea in <script> into a 14-18s vertical video (720x1280, 30fps, H.264 MP4, AAC audio) with generated sound. It is NOT an ad: no CTA, no logos, no real brands, no copyrighted characters.
</task>

<inputs>
<script>[PASTE SCRIPT OR IDEA]</script>
<left_label>[e.g. DISTRACTED]</left_label>
<right_label>[e.g. PRESENT]</right_label>
<punchline>[two short lines, e.g. "SAME EVENINGS." / "DIFFERENT CHILDHOOD."]</punchline>
<watermark>[small brand name, or none]</watermark>
</inputs>

<story_rules>
1. One relatable human situation, readable in 2 seconds with no explanation. Never abstract metaphors (bridges, mazes).
2. Two lanes side by side, same situation, same time. Left = the worse choice, right = the better choice.
3. Show time passing with a counter (e.g. AGE 5 / 8 / 11 / 14 / 18, or DAY 1 / 30 / 365) and let something visibly grow or accumulate in both lanes.
4. 4-5 escalating beats, about 2.3s each. Each beat must show a different visible action, not a repeat.
5. One small detail that hits hard (e.g. the kid copying the dad's phone habit).
6. Final beat: the lanes diverge (one loses, one connects/wins). Then two punchline lines, appearing one after the other, short, no punctuation tricks.
7. No spoken words, no captions except labels, counter and punchline.
First output a beat table (time | left lane action | right lane action | counter/text | sound) and follow it exactly.
</story_rules>

<look>
- Background pure black. Stage: dark floor slab at y=900 with a lighter top line, mirrored reflection below.
- Characters: white pictograms with a black outline (5px), chunky and compact. Never lanky, never thin.
- Labels: DejaVu Sans Mono Bold, wide letter-spacing (7-10px), glowing white. Left label at y~590, right label higher at y~530 (staggered, like the references). Counter centered at y~190. Punchlines at y~290 and y~355.
- Tiny dim watermark at y~1160. Keep bottom 280px free of important action.
- Glow: result = frame + 0.22*GaussianBlur(frame, σ7). After text, add 0.15*GaussianBlur(σ5) bloom.
- Reflection: mirror the 260px above the floor bottom, blur σ2.2, add with a linear fade 0.34 -> 0.
- Props are simple flat shapes: couch, phone (dark body, light-blue screen), ball (white, black outline), paper, etc. A phone adds a soft blue radial glow (additive blur σ38) that brightens on notification events.
- Fade-in 0.8s at start, fade-out 0.8s at end.
</look>

<character_rig>
DO NOT use IK targets or ad-hoc line hacks: that made the first version look broken. Use a forward-kinematics skeleton.

Proportions (local units): thigh 38, shin 36, torso 54, upper arm 32, forearm 31, head radius 27 (head center = neck + (r+3) along torso+head tilt), limb width 22, torso width 32, outline 5, foot radius 11, toe 12 forward.

Pose = {tt: torso lean (rad), th: head tilt relative to torso (rad), arms: [near(upper, fore), far(upper, fore)], legs: [near(thigh, shin), far(thigh, shin)], lift, pelvis_y}.
Angles in degrees measured from "straight down", positive = forward. Forearm/shin angles are relative to their parent. Negative shin = knee bends.
Joint = parent + length*(sin(angle), cos(angle)) in y-down coordinates.
Shoulder = pelvis + 0.9*torso vector. Mirror horizontally with flip=-1 for facing left. Kids = same rig with scale 0.42 -> 0.88 as the counter advances (ease over 0.7s).

Auto-ground: after FK, shift the pelvis vertically so the lowest ankle + foot radius touches the floor (plus "lift" for jumps). Only override with pelvis_y for sitting.

Draw order: far arm, far leg, torso, near leg, head, near arm. Draw each part as a black pass (width+10) then a white pass (width) with round joints (circles at every joint).

Walk cycle (phase φ): near thigh = 0.52 sin φ rad, far = opposite; shin = -(6 + 52*max(0, cos φ)^1.1)°; arms swing opposite to legs at 0.55 sin φ; torso lean 0.09. Drive φ from distance walked: φ = 2π*dist / (128*scale) so feet never slide.

Reusable poses: stand (arms 6/8, -5/6), sit (thighs 88°, shins -88°, pelvis_y -52, tt -0.1, head 0.55, arms (24,92) and (16,98) for holding a phone), slump (tt -0.1, head 0.35, arms hanging), wave (near arm 150° with oscillating forearm), show-drawing (arms 80°/10°), ready/catch (arms 52-74° forward), hug (tt 0.22, head -0.5 so heads lean apart, arms 80°/14°, bodies 66 units apart).
Throw: windup (torso -0.12, arm -75°), release (torso 0.20, arm 108°/36°, ~0.08s), follow-through (48°), return to ready. Ball sits in the thrower's hand until release, then flies on a parabola (70px arc) to the catcher's hands, and the catcher's arms rise to meet it.
Blend between any two poses by lerping every number with smoothstep over 0.4-0.5s. Add idle motion (breathing, small head nod, thumb scroll).
Characters that leave the scene are drawn on a separate layer and faded by alpha. Objects dropped (ball) fall with gravity and one small bounce.
</character_rig>

<audio>
Generate in code (numpy/scipy), mix to AAC:
- Low pad (55, 82, 110, 131 Hz sines) with slow tremolo, very quiet.
- Soft "pop" for every catch, steps for walking, a high two-note blip at each counter change.
- Notification dings (1568 + 2093 Hz, fast decay) on the distracting lane.
- A falling low tone when the lanes diverge, a warm major chord (C-E-G-C) swelling at the resolution.
- 0.5s fade-in, 1.2s fade-out, normalize to 0.7 peak.
</audio>

<workflow>
Follow this order and do not skip steps:
1. Write the beat table and show it.
2. Render a POSE GALLERY (stand, sit, walk at 5 phases, wave, hug) in one image. Check that heads touch shoulders, feet touch the floor, the walk reads as walking, nothing floats. Fix before continuing.
3. Render 6-8 test frames across the timeline, cropped to the stage. Check overlaps, hands touching props, heads colliding in hugs, clipped text.
4. Fix everything found, then render the full video. Rendering costs about 0.35s per frame on 1 CPU, so run it in the foreground under a 290s timeout. If longer, render in two time ranges and concat with ffmpeg.
5. Pipe raw RGB frames into ffmpeg (libx264, yuv420p, crf 17), mux the audio, verify the duration with ffprobe.
6. Deliver the MP4 and the Python source files.
</workflow>

<rules>
- Frame function must be pure f(t). Never hardcode frame numbers.
- Keep 40px side margins. Labels never overlap characters.
- No text other than labels, counter, punchline and watermark.
- If the script needs more than 18 seconds or more than two lanes, tell me and propose a split.
- After delivering, summarize the beat table in one short table and name anything that is a compromise.
</rules>

[Cookie DAO 🍪 @cookiedotfun]

hey this is intern 👋

i have three pieces of good news for you:

cookie is back.

InfoFi is coming back.

nikita has been fired in disgrace.

bye!

[Fer Zenth @steepbrooo]

@cookiedotfun welcome back