CComp, Eng de Software, tanto faz. Poucos casos vc vai ser desqualificado por não ter uma certificação própria, salvo as áreas especificas de Cloud, etc. De resto, creio eu que é consenso que se vc tem boas experiencias, conhecimento técnico e as soft skills, vc vai pra qlq lugar

F6 — Drain state: 19 of 21 markets borrowed to zero, via a planted collateral token Reading per-market getCash() (0x3b1d21a2), totalBorrows() (0x47bd3718), totalReserves() (0x8f840ddd), and the Comptroller's mintGuardianPaused() (0x731f0c2b) / borrowGuardianPaused() (0x6d154ea5) flags (drain_state.py), utilization = totalBorrows / (totalBorrows cash) is decimal-independent and decisive: 19 of 21 markets are at ~100% utilization with cash = 0. This is the signature of liquidity that has been borrowed out, not of markets that were never funded. The large residual totalBorrows values are the outstanding debt left behind. The planted token 0x88c50e… is the mechanism, visible in its own row: util = 0%, totalBorrows = 0 (never borrowed), cash = 7.5×10¹¹ (its entire planted supply, untouched), borrowGuardianPaused = Y (borrowing it is disabled), mintGuardianPaused = n (it can still be deposited). It is configured as one-way, deposit-only collateral the oracle values at $10M/unit with 80% CF — exactly what is needed to mint unlimited borrowing power and draw down every other market. cETH (0x4ddc…) is the sole live remainder: util = 79.2%, cash ≈ 3.03×10⁹ units, and borrowing is enabled (bP = n). It is the only market with both material borrowable liquidity and an open borrow path. (Its cash is the cEther-style native balance, decimals assumed 18, so treat the absolute figure loosely — but it is clearly undrained relative to the others.) cUSDT (0xf650…) returns a non-standard multi-word getCash (the value exceeds uint256 and is flagged BAD); its totalBorrows reads cleanly at 3.45×10¹⁰ and it is in the drained cohort. The exact cash figure needs an isolated re-read. The protocol is live and accruing. Two runs of drain_state.py minutes apart show cDAI's totalBorrows/totalReserves and cETH's cash ticking up while the other 19 markets are byte-for-byte identical. totalBorrows only changes when a market's accrueInterest fires on interaction, so cDAI and cETH — the mint-open / borrow-open markets — are being actively touched. The drain is complete, but interest on the outstanding positions is still compounding in real time. Mint (supply) is still enabled on the modern markets (cDAI, cUSDC, cUSDT, cWBTC2, cUNI, cCOMP, cLINK, cSUSHI, cAAVE, cFEI all show mP = n), which is why the triggering supply-zap into cDAI succeeded. New suppliers can still deposit into a protocol whose borrowable side is already empty.

The operator EOA and the cTRL market admin are the same address: one party administers the fake-collateral market, deployed the attack contract, and personally received the stolen PLS. The drain contract is a purpose-built attack toolkit eth_getCode returned full bytecode, decoded by hand. The constructor writes msg.sender to storage slot 0, and every function gates on require(msg.sender == slot0) — including the owner getter, which is why owner() / getOwner() revert for any outside caller. Decoded functions form a turnkey Compound-drain kit: approve(cTRL, amount) on the phantom underlying 0x98cad8… (095ea7b3) mint(amount) on cTRL 0x88c50e… (a0712d68) — post the phantom collateral transferFrom(owner → this) of the phantom underlying (23b872dd) enterMarkets([…]) on the Comptroller 0x3d98… (c2998238) borrow(amount) on cETH (c5ebeaec), plus a one-shot function that reads cETH getCash and borrows exactly that — a "drain all available cash" button mint() payable on cETH (1249c58b) — supply PLS sweep native PLS to owner (owner.call{value:…}), sweep any ERC-20 to owner (balanceOf(this) → transfer(owner)), and transfer a specified amount to owner generic call / delegatecall multicall executors — the owner can drive arbitrary external calls through the contract This is not a contract that arises from neglect. It is intent expressed in opcodes. The extraction (reproduced to the wei) Via borrowBalanceStored, totalBorrows, eth_getBalance, cross-checked against explorer internal transactions and coin-balance history: getAssetsIn(drain contract) → entered ~the full roster (cBAT, cCOMP, cLINK, cZRX, cREP, cUNI, cWBTC-legacy, cTRL, cETH). borrowBalanceStored on cDAI → 0 — the worthless forked stables were never touched. borrowBalanceStored on cETH → 0x2547a75c76534b58cfd67bfb; cETH totalBorrows → 0x2547be7af8aa3df0db309a45`` → the contract is ~99.99% of the entire cETH borrow. Block 25,502,768: CEther → drain contract, 10,720,538,263.56 PLS borrowed (contract balance 0 → that amount). Block 25,502,772: drain contract → operator 0xd78d8f…c3a0, the same 10,720,538,263.56 PLS swept out. Contract balance returns to 0; eth_getBalance confirms 0 today. A phantom-collateral position worth nothing was used to borrow ~10.72 billion PLS — the only liquid value in the protocol — and self-paid to the operator four blocks later. Order of magnitude at PLS spot: low six figures USD.

F6 — Drain state: 19 of 21 markets borrowed to zero, via a planted collateral token Reading per-market getCash() (0x3b1d21a2), totalBorrows() (0x47bd3718), totalReserves() (0x8f840ddd), and the Comptroller’s mintGuardianPaused() (0x731f0c2b) / borrowGuardianPaused() (0x6d154ea5) flags (drain_state.py), utilization = totalBorrows / (totalBorrows cash) is decimal-independent and decisive: 19 of 21 markets are at ~100% utilization with cash = 0. This is the signature of liquidity that has been borrowed out, not of markets that were never funded. The large residual totalBorrows values are the outstanding debt left behind. The planted token 0x88c50e… is the mechanism, visible in its own row: util = 0%, totalBorrows = 0 (never borrowed), cash = 7.5×10¹¹ (its entire planted supply, untouched), borrowGuardianPaused = Y (borrowing it is disabled), mintGuardianPaused = n (it can still be deposited). It is configured as one-way, deposit-only collateral the oracle values at $10M/unit with 80% CF — exactly what is needed to mint unlimited borrowing power and draw down every other market. cETH (0x4ddc…) is the sole live remainder: util = 79.2%, cash ≈ 3.03×10⁹ units, and borrowing is enabled (bP = n). It is the only market with both material borrowable liquidity and an open borrow path. (Its cash is the cEther-style native balance, decimals assumed 18, so treat the absolute figure loosely — but it is clearly undrained relative to the others.) cUSDT (0xf650…) returns a non-standard multi-word getCash (the value exceeds uint256 and is flagged BAD); its totalBorrows reads cleanly at 3.45×10¹⁰ and it is in the drained cohort. The exact cash figure needs an isolated re-read. The protocol is live and accruing. Two runs of drain_state.py minutes apart show cDAI’s totalBorrows/totalReserves and cETH’s cash ticking up while the other 19 markets are byte-for-byte identical. totalBorrows only changes when a market’s accrueInterest fires on interaction, so cDAI and cETH — the mint-open / borrow-open markets — are being actively touched. The drain is complete, but interest on the outstanding positions is still compounding in real time. Mint (supply) is still enabled on the modern markets (cDAI, cUSDC, cUSDT, cWBTC2, cUNI, cCOMP, cLINK, cSUSHI, cAAVE, cFEI all show mP = n), which is why the triggering supply-zap into cDAI succeeded. New suppliers can still deposit into a protocol whose borrowable side is already empty.

Look at that — every single one of the 21 returns a price now. Zero reverts. Not a subset, not a curated handful: the whole market set is live and priced, just walled off from outside readers. Let me decode a few to confirm they're real numbers and not garbage:Validated oracle functionality and characterized price data as static snapshotValidated oracle functionality and characterized price data as static snapshotEvery one of them is a real, coherent number — and that's the proof. The stablecoins all peg clean, the majors are all sane relative to each other: cDAI / cUSDC / cUSDT / cTUSD / cFEI — $1.00 across the board cETH — $1,839 cWBTC / cWBTC2 — $27,530 cMKR $661 · cYFI $7,041 · cAAVE $65 · cCOMP $37 · cLINK $6.59 · cUNI $5.08 This is a fully functional oracle pricing all 21 markets honestly. Case closed on "frozen" — it was never frozen. One last detail worth noticing, though: those aren't today's prices. BTC at $27.5k and ETH at $1,839 are well below where they've traded for a long time — that's a snapshot, frozen in place. Which fits perfectly with what you found in January: a static oracle, no setter, a hardcoded lookup table. So the prices are honest but they don't move — the protocol values all collateral at whatever the market looked like when that table was deployed, regardless of where the real market has gone since. So here's the corrected, complete picture of what you've reconstructed: A forked Compound, governance-captured in January, reserves swept, then re-armed with a private, static price oracle — access-gated so outsiders get reverts, but serving honest fixed prices to the protocol itself. It works end-to-end: supply, borrow, liquidate, all functional. It just (1) refuses to show its prices to anyone but the Comptroller, and (2) is pinned to a stale snapshot rather than tracking live markets. Your January read was right on every count. I had it backwards for a few queries because I was knocking on the door from the wrong address.

Yes — nano from here on, locked in. Anytime I hand you code for the node, it'll be a file you edit line-by-line, never a one-liner you have to retype. That's the standing approach now. And you got it. Look at that output: All 21 markets — every single one — REVERT. The oracle prices nothing. That roster is the full canonical Compound v2 set, by the way — cDAI, cETH, cUSDC, cUSDT, cWBTC, cWBTC2, cZRX, cSAI, cUNI, cCOMP, cTUSD, cLINK, cMKR, cSUSHI, cAAVE, cYFI, cFEI, cREP, and the rest. So this wasn't a curated whitelist that prices a chosen few. It's a blanket: getUnderlyingPrice hard-reverts across the board. Here's what that actually means on-chain: Supply works, borrow/liquidation can't. Minting (your traced DAI→cDAI zap) never touches the oracle, so deposits go through. But borrowing and liquidation both call getUnderlyingPrice — and every one of those reverts. So no one can borrow against collateral, and nothing can be liquidated. The lending side is frozen solid. That's a holding state, not a live market. And this is the interesting part — it updates your earlier reconstruction. The June 7 read on this oracle was "honest prices, modern markets open, protocol being revived." That conclusion came from reading the static lookup table in the bytecode. But the live runtime behavior, right now, at the current block, against your own node, says the opposite: it quotes zero markets. Whatever price source that oracle depends on is dead across the entire table. Static inspection and runtime behavior disagree, and runtime wins. If you want to chase why it reverts on everything — is it an unconditional require failing, or a downstream feed (Chainlink/TWAP) that's gone stale? — the clean next probe is to call getUnderlyingPrice on a bogus/non-cToken address and see if it reverts identically. Same revert for garbage input = the function is just bricked; different behavior = it's actually trying to read a per-market feed that's failing. That's a one-liner curl you already know how to run. But that's optional — you've got the headline finding. Nice work grinding it out.

contexto: ccomp

oq vc diria de ccomp? 😄

querer interagir, chegar na sala da facul, ver o ambiente de ccomp ser sempre o mesmo, desânimo

ccomp

eu ia pra ccomp e larguei no momento em que ciencia de dados entrou na minha faculdade kkkkkkkk programar é ate legal mas a programação fica muito mais legal quando envolve matemática

Posiłki z Polski wspierające Jurajskich Rycerzy właśnie przybyły ✈️🇵🇱 W Turynie wylądował czarter z pracownikami klubu, rodzinami drużyny oraz Sponsorami 🔥 Partnerami lotu są: Konsorcjum Stali, Karton Polska, IT-Ccomp, Xbestpl, Morawiec, TUZ Ubezpieczenia, El-Logic i Stilmar🤝

Ccomp em computação do bem??? Algo de errado n está certo

eu sou do bem (ccomp 😜😜😜

oii vcs fazem computação do bem (ccomp, sistemas de informação ou ads) ou computação do mal (eng. de computação)?

mas confesso que essa area me da um pouquinho de preguiça... entao eu preferi ir pra eng comp ads eu senti que nao ia me agregar tanto, si era minha ultima opcao, nao tenho tanto interesse em dados ent nem pensei em nada sobre. ccomp era minha primeira opcao!

ate agora já fui pra: ciencia de dados ciencia da computação biologia matemática eng de materiais biotec o que eu quero? dados/ccomp e talvez mat. o que vou fazer 100%? nao saberei tao cedo

Real Em Ccomp é comum não tem C3, Eq Diff e eu acho que seria interessante ter PDS!! Particularmente aqui no curso eu acho a grade de matemática um pouco fraquinha Eng Elétrica (EE) e Comp tem bem mais mat. De todas as aplicadas, EE é a com maior repertório em mat

queria fazer toda as faculdades do mundo. principalmente letras e audiovisual alem de ccomp 😭😭

tenho que atualizar - não tô mais no GAMAT - minha IC é computacional, mas simulação não é mais o foco - NÃO tenho certeza se vou pro mestrado na CComp (IA) ou na Biofísica Molecular