Curve DAO Seeks New Risk Monitor as Bid Flags $265M crvUSD Exposure

Curve DAO Splits crvUSD and LlamaLend Risk Work

The DAO divided the mandate into two areas. The first covers crvUSD mint markets, PegKeepers and relevant DAO credit lines. The second covers isolated LlamaLend markets and the LLAMMA infrastructure that supports Curve’s soft-liquidation design.

For the crvUSD scope, the call asks a provider to assess new markets before votes, evaluate collateral and oracle dependencies, monitor liquidity and abnormal activity, and recommend parameter changes when needed. For LlamaLend, it asks for comparable collateral, liquidity, liquidation and solvency work across lending markets.

That separation is practical rather than cosmetic. A PegKeeper helps crvUSD trade near its intended value, while an isolated lending market has its own collateral, oracle and liquidation mechanics. Both can create losses if stress is missed, but the warning signs and response tools can differ.

The call says the selected team should make reusable tools, models, documentation and monitoring assets available to the DAO after the mandate. That handoff requirement is material after a provider departure: a dashboard is less useful if Curve cannot operate, inspect or transition it when the contract ends.

The standards are familiar across DeFi. A recent Bonzo Lend oracle failure showed how a price-verification weakness can turn an input problem into outsized borrowing, while Aztec’s legacy-contract exploit illustrated a different code-path risk. Neither event is a direct comparison with Curve’s systems.

Xerberus Bid Puts a $265M Credit Line in Focus

Xerberus submitted a July 19 proposal for the LlamaLend and isolated-markets scope. The firm says it reconstructed Curve markets from public on-chain data and found that about $265 million was drawn from a $1 billion authorized Yield Basis credit line, which it describes as Curve’s largest single crvUSD exposure.

The proposal says that exposure sits outside every LLAMMA-era monitoring surface it reviewed. That is a material allegation, but it remains the view of an applicant seeking the mandate. Curve has not published a confirmation of the measurement, a rebuttal or a decision to change the credit line.

The same bid says 79% of borrowed value in its reconstructed LlamaLend book was concentrated in three stable-loop markets, where yield-bearing stablecoins are used as collateral to borrow other stablecoins at high loan-to-value ratios. It also says a larger market can face an exit-capacity problem if collateral cannot be sold quickly enough during liquidations.

Its proposed service would monitor active LlamaLend markets on a 12-hour cadence, publish a monthly digest and deliver a written escalation within 12 hours of a critical trigger. Xerberus seeks $100,000 a year in CRV, saying all received tokens would be locked as veCRV, with a three-month public checkpoint.

Pharos Proposal Sets $350,000 crvUSD Monitoring Ask

Pharos Watch submitted a separate July 18 proposal focused on crvUSD mint markets. It says its current system already provides peg, safety and liquidity scores and real-time depeg alerts, while several parts of the requested mandate would be built or formalized after an award.

The proposal seeks $350,000 a year, split between $100,000 in stablecoins and $250,000 in CRV. It proposes a three-month trial and says all CRV received under the requested split would be maintained as veCRV. Its target for a critical human escalation is four hours, compared with the 12-hour written-escalation commitment in the Xerberus proposal.

Pharos says its scope would include every active crvUSD market and PegKeeper pool, pre-vote reports, parameter recommendations and public reporting. It does not bid for the isolated LlamaLend scope, arguing that those markets require a separate scenario-testing capability.

The public bids do not establish a winner, a deadline for a vote or an agreed price. They do show that the choice includes trade-offs among scope, response time, deployment readiness, ownership and compensation, rather than a simple replacement of one vendor with another.

The Crypto Fear and Greed Index was at 34, or Fear, when checked July 20. The Bitcoin-centered gauge does not measure Curve-specific risk, but it adds market context to a governance process focused on stablecoin and lending safeguards.

What comes next is a governance decision, not an automatic contract award. The evidence to watch is whether Curve contributors publish further bids, define a review or vote timetable, verify or challenge the exposure figures and explain which monitoring tools will remain usable by the DAO if another provider leaves.