ThroneGame18™📈

When you see multiple versions of something pop up in a week… You’re not early — you’re witnessing signal. Since the last Percolator update, the timeline has been flooded: • Fake launches claiming to be the “real one” • People chasing narratives • Noise everywhere And then reality hits: “Anyone that calls any token or project mine is trying to deceive you.” ~toly Clear. Now look at what actually matters: • Official account going hard with threads + updates • Followers growing fast • Devnet already active with 60+ markets • Contributors getting paid to test + build • Spaces being hosted, more coming • Direct interaction happening at the protocol level No shortcuts. No fake claims. Just builders shipping and the ecosystem forming around it. 🚢 🚢 I’ve been heads down building @solanaOilFac the past few days, but Percolator is still a big focus. Still early to see it. $Percolator 🦞☕️

🧵🪡 1/4: Migrated all three Percolator programs to pinocchio token this week. Big shoutout to the anza-xyz team for building and maintaining this. pinocchio token replaces the standard spl token CPI with a leaner implementation purpose built for Solana programs. SIMD-0266 passed governance The numbers: Token Transfer drops from 2,400 CU to 131 CU. MintTo drops similarly. 18x cheaper at the instruction level.

Don’t trust, verify. percolatorlaunch.com/developers

here is what has been shipping over the last few days.🚢 The markets list is now accurate: The app was showing 168 markets when only 122 were actually active, and the homepage was showing different numbers to the stats page. All of that is cleaned up now. What you see is what is real. Zombie markets are handled properly: Some markets had no real liquidity or were inactive but were still showing up in the list with bad prices. These are now filtered out correctly. The markets you see are live. Prices are displaying correctly: One featured market was showing a $10 billion price because of a decimal conversion bug = Fixed. All prices on the homepage and market cards are now accurate. The faucet works reliably: The devnet faucet (the button that gives you test tokens) had some issues returning the wrong error codes and occasionally failing = All patched up. Getting test tokens to try the app is smooth now. Security updates: 15 dependency vulnerabilities fixed in one go by updating key packages. Routine but important before mainnet. Funding and open interest history is clean: The charts and stats tracking trading activity over time were including some phantom data from inactive markets = Patched and verified. Long live the Percolator ☕️☕️

SOL faucet was returning 500 errors on rate limit instead of the correct 429 -> fixed. Mint authority mismatch on the faucet also resolved. Oracle unavailable banner now has a proper early return guard so it doesn't render in states it shouldn't.

The Percolator app has come a long way since we submitted it. Connect a wallet with no funds and it handles everything automatically -> detects your balance, drops SOL and USDC, and deposits straight into your account. You go from zero to ready to trade in seconds. Market creation works properly now. The full 5 step wizard flow has been tested end to end on devnet. If a step fails mid way, you can resume from where you left off without losing your progress or your SOL. Balance checks before you submit so you never hit a wall at step 4. The dashboard shows accurate numbers. OI, volume, and stats are all pulling from the same source now with proper decimal handling. What you see is what's actually on chain. Markets with no live oracle now show the last known price clearly labeled, so you always have a reference point. Token Factory is working again after a couple of broken states got ironed out. The app is getting closer to something you’d actually want to use everyday. Percolator App preview:

Would love to some feedback on what will likely be a big change, github.com/aeyakovenko/pe…

🧵🪡 1/5: LP vaults vs liquidity pools: what’s the difference? If you have provided liquidity on Uniswap or Raydium, you understand liquidity pools. Perp LP vaults work very differently.

A lot has been being fixed, tightened and verified quietly in the background. These are “boring” things that matter, you can always track the GitHub directly: github.com/dcccrypto/perc… Or go to the website, and check the developer section for all repos and contributions: percolatorlaunch.com/developers This is what this stage looks like: Stats layer is now reliable: The OI aggregation pipeline had a decimal mismatch -> raw u64 on chain values were passing through without dividing by token decimals, producing trillion dollar figures in the DB. Migration 047 patched the root cause, phantom OI cleared across 25+ markets, and `StatsCollector` now has full boundary test coverage for the `vault=0` and dust-vault edge cases that were triggering the phantom values. EngineHealthCard sanitization: The health tab was reading `cTot`, `pnlPosTot`, `netLpPos`, and `lpSumAbs` directly from on-chain u64 bytes -> showing `8.2e18` where it should show `0.00`. `sanitizeOnChainValue()` is now applied consistently across all engine fields. Funding route hardening: `/funding/global` and `/funding/:slab/history` were throwing 500s on any DB error. Applied graceful-degrade pattern with proper error handling and regression tests. All three funding endpoints now return clean responses under failure conditions. Rate limiting is distributed: In memory rate limiting was per-serverless-instance -> not enforcing across concurrent requests at scale. Migrated to Upstash Redis-backed rate limiting on Vercel. Production grade now.

Most people who trade perpetual futures never stop to ask one important question - Who actually decides the price you’re trading at? On a normal exchange you see numbers moving up and down SOL is $70, then $75, then $82... But behind those numbers is a system constantly calculating what the fair market price should be That system is called the pricing engine Think of it like a referee in a football match. The referee is not playing the game, but they make sure the rules are followed so nobody cheats. In trading, the pricing engine plays that role. If the pricing system is weak, traders can manipulate prices with small trades For example, imagine a trader places a tiny order just to push the price slightly higher, triggering liquidations for other traders. This has happened many times across crypto. That’s why exchanges spend a lot of time designing pricing engines carefully. What’s interesting about Percolator is that the project is building its pricing system with multiple layers of checks instead of relying on one signal So instead of blindly trusting one number, the system constantly compares different pieces of information before deciding the price For traders this matters more than people think Because a strong pricing engine means fewer unfair liquidations, fewer manipulation attempts, and a much more stable market And in trading, stability is what separates serious infrastructure from short-lived experiments Watch out for more $Percolator educational contents

As Percolator gets closer to mainnet, it’s worth understanding what the protocol is actually trying to solve. Most perpetual exchanges decide which markets exist. Percolator flips that model. Anyone can create a perpetual market for any Solana token. No listings. No approvals. No gatekeepers. Each market can have its own liquidity, leverage, and trading activity, all fully on-chain. That means the number of markets isn’t limited by an exchange… It’s limited only by what the community wants to trade. If this model works at scale, it could unlock a completely new layer of trading infrastructure on Solana. Mainnet will be an interesting experiment to watch. $Percolator ☕️🦞

🧵🪡1/6: Most people don’t understand what a perp dex is or the fancy tech language behind Percolator. Part of our job is teaching and helping people understand that so they know what they’re investing into. So let’s talk about what actually happens when you get liquidated.

🧵🪡 1/6: why do funding rates exist? Every time you hold a perp position overnight you pay or receive funding. Most traders treat it as a tax; Here is what it actually is:

Slopped together this video to keep yall bullish on $percolator! The future is free!! Load tf up!! @Percolator_ct Ca: 8PzFWyLpCVEmbZmVJcaRTU5r69XKJx1rd7YGpWvnpump #solana #pumpfun $sol #crypto #100x #10x #hackathon #percolator

🧵🪡 1/7: Most people trading perps have no idea what sets the price they trade at, let Percolator explain exactly how it works:

We audited our own Kani proofs. Found gaps & fixed them. 189 proofs sounds solid. But a security audit this week found that our liquidation proofs -> the highest risk property in the entire codebase were all concrete. Hardcoded inputs, no symbolic coverage. Technically passing. Not actually proving anything meaningful about arbitrary states. So we upgraded them. The new proofs use `kani::any::()` — the engine state is fully symbolic. The prover checks the liquidation invariant across every possible input, not just the ones we thought to write. That is the difference between a unit test and a formal proof. Also caught: 5 proofs had `kani::assert()` where they needed `kani::assume()` — using assert as a precondition silently weakens the proof by letting Kani skip cases instead of constraining the input space. Fixed across all 5. And one vacuous proof that was designed to fail Kani (a negative test) but was tagged `#[kani::proof]` which meant it was running in CI and breaking the build for the wrong reason. Removed the attribute, kept the logic as a regular test. Three fixes. But the liquidation one matters most. The invariant that prevents bad debt from propagating through the engine is now proven over the full symbolic state space. Not just the cases we anticipated. This is what formal verification is actually for. Percolator stays building, and we are consistently working our magic to improve and build in every aspect. ☕️☕️☕️

🧵🪡 1/8: Most perp protocols are still figuring out their oracle setup. Percolator just shipped quadratic funding, self funding keepers, elastic shared vaults, adaptive creation deposits, and creator stake locking. All in 24 hours. All on Solana.

We are on track to submitting the Percolator app to the Solana Mobile Hackathon, Stay tuned for the official announcement.. TODAY.

We are prepping for one of our biggest updates to day. $Percolator is about to get more real, and one more step towards being ready for main net and testing in the later stages. Stay tuned. 8PzFWyLpCVEmbZmVJcaRTU5r69XKJx1rd7YGpWvnpump

Big push on the Percolator pricing engine last night: We shipped PERC-117, -118, and -119 back-to-back -> full on chain pricing stack complete: - PERC-117: Pyth oracle CPI live -> pulls real time prices directly on chain (UpdateMarkPrice, 83 tests passing) - PERC-118: Hyperp EMA blend for mark price = blend(oracle_price, impact_mid_price) -> exactly how a permissionless perp should be done (PRs #21 + #57, security pass done) - PERC-119: Full funding rate engine wired -> premium computation, accrual, keeper crank all confirmed Pipeline now: Pyth CPI -> EMA-blended mark price -> clamped funding rate -> Core permissionless perp pricing loop = done Kani side: moved heavy fee_monotone_in_bps proof to nightly CI (PR #60) so PR checks stay fast. #16/#20/#21 queued & green. Security: CSP headers tightened (no more vercel.app wildcard), unsafe-eval + frame ancestors fixed, several issues closed. Nothing blocking. Devnet rock solid -> oracle keeper pushing BTC ~$67-68k every 3s, CI 5/5, deploys flowing again. Percolator perps infrastructure coming together fast.