Ajna v2 Users Told to Exit After $775K Liquidation Exploit

Ajna v2 Exit Notice Follows Seven-Pool Drain

Ajna’s public-facing materials characterize the protocol as a peer-to-peer, non-custodial lending system that operates without external price feeds. That architecture is relevant to the reported attack: the public accounts point to the protocol’s own liquidation and accounting path, rather than a compromised wallet, a stolen key or a manipulated external price feed.

Ajna’s official reference materials define a quote token as the ERC-20 asset lenders deposit into price buckets. The reported user instruction to withdraw quote tokens and repay loans is specific to that design. It does not establish that every user has the same balance status or that every withdrawal will complete, so lenders and borrowers need to check the live interface and official project channels before signing a transaction.

The loss-impact review found no public evidence, as of this article’s 22:10 Casablanca-time review, of a larger confirmed theft, a separate wave of attacks, an affected-wallet total, recovered assets or an official final incident report. DefiLlama says Ajna’s tracked total value locked was about $232,779 and down 73% over 30 days, but that dashboard statistic should not be treated as a substitute for a forensic balance sheet.

The episode illustrates a harder trade-off in immutable lending software. Immutability can eliminate discretionary upgrade control, but it can also remove the fast containment tool users may expect after an exploit. Recent Cosmos EVM losses across six chains involved a different shared-software issue; both cases, however, show why public scope and remediation details matter as much as a headline loss number.

Liquidation Logic, Not an Oracle or Wallet Breach

The available reports describe a liquidation-accounting manipulation. Ajna’s system uses price buckets and internal liquidation mechanics instead of a conventional outside oracle. Reporting on a cbETH pool transaction said an attacker combined the protocol’s liquidation functions with flash-loaned assets and extracted collateral while paying comparatively little quote token, but Ajna has not independently confirmed that reconstruction.

That distinction should remain precise. There is no public basis yet to say that Ethereum itself was compromised, that user seed phrases were exposed, or that the event was caused by a general failure of every oracle-free lending model. There is also no basis to say the listed $775,400 has been fully recovered or that losses have stopped permanently.

The first practical checkpoint is Ajna’s own post-mortem: it would need to establish the vulnerable condition, the complete transaction set, user effects and whether any alternative interface or migration route is available. Until then, the protocol’s exposure notice and the independent incident ledger are more informative than an unverified claim that the contracts have been repaired.

For a broader risk comparison, the difference between a smart-contract exploit and a platform operational response is also visible in the MANTRA incident post-mortem, which later documented its transaction trail and containment details. Ajna had not done so by publication.

No Recovery or Patch Timeline Is Public Yet

Ajna’s brand and product site emphasize that the protocol has no governance and no price feeds. Those are design characteristics, not a post-exploit assurance. The project has not disclosed a contract-level pause, replacement deployment, reimbursement program, recovery agreement or target date for a technical update.

The projected reader action is narrow: the project’s reported notice says users should exit, repay and cease interaction; users should verify any subsequent instructions through Ajna’s official channels and treat unsolicited links or transaction requests as potential phishing. A security alert is not a reason to rush into an unverified workaround.

Ether traded near $2,467.30 at the Aug. 30 public market-data reading, versus a $1,844.61 close on Aug. 1, according to CoinGecko’s historical page. The market move does not determine the exploit total, but it affects the dollar value assigned to Ether-based collateral and any eventual recovery.

Broader crypto sentiment remained in Greed territory. Alternative.me’s Crypto Fear and Greed Index read 69 on Aug. 30, after a 68 reading the prior day. That market-wide gauge does not measure Ajna’s condition or resolve the outstanding user-fund questions.

Ajna’s estimated $775,400 drain is small beside the year’s largest exploits, but the user impact is immediate because the protocol’s reported containment method is exit rather than a central pause. The next facts worth watching are a project post-mortem, on-chain loss reconciliation, affected-wallet and pool details, and evidence of any recovery or safer migration path.