Problem
Ethereum’s consensus layer has a structural blind spot: individual protocols optimize for their own UX while creating systemic hidden risk. The clearest example is Lido’s oracle: when MEV rewards exceed slashing losses, the rebase is positive, Bunker Mode stays off, and withdrawals finalize at 1:1 — even while slashing debt accumulates silently among remaining stakers. This is Algorithmic Moral Hazard — protocols won’t self-regulate at the cost of their own liquidity.
No on-chain coordination layer exists to signal this risk before it becomes a liquidation cascade.
Solution
QUEST is a macroprudential oracle AVS — the on-chain equivalent of a Bank for International Settlements (BIS) for Ethereum. It computes a Grey Zone Score (gross_slashing_loss / (cl_rewards + burned_eth)) from live Beacon chain data and publishes it on-chain as a PMC signal (θ) every epoch (~384s).
DeFi protocols implement IERC8004QuestAware to receive these signals and adjust defensively before systemic risk materializes — without coercion, purely opt-in.
What’s Already Built
| Component | Status |
|---|---|
| Risk engine (Python, Beacon REST + Alchemy) |  Running on Cloud Run |
| QUESTCore.sol — on-chain oracle | Sepolia 0xE81C6B16… |
| QUESTAwareProtocol.sol — ERC-8004 reference impl | Sepolia 0xD693C783… |
| AVS node (Go) — submits epoch metrics on-chain every ~384s | Running on Cloud Run |
| Live dashboard — Grey Zone Score + 3-layer decentralized storage | quest-orcin-sigma.vercel.app |
Every epoch snapshot is persisted in Firestore + IPFS (Pinata) + Filecoin (Lighthouse) in parallel. Each epoch is independently verifiable at /epoch/{n}.
Current architecture: trusted oracle v1 (single ECDSA operator). The grant funds the upgrade to a proper EigenLayer AVS with a BLS operator set and cryptoeconomic security.
Milestone Plan
M1 — EigenLayer SDK integration (8 weeks)
Port quest-avs-node from trusted operator to EigenLayer AVS SDK. Implement ServiceManager, BLS key registration, operator slashing conditions.
M2 — Operator testnet (4 weeks)
Launch with 3–5 independent operators on Sepolia. Validate consensus on Grey Zone Score across operators.
M3 — Mainnet AVS (4 weeks)
Deploy QUESTServiceManager on mainnet. Open operator registration. Publish ERC-8033 signal certified by restaked ETH.
Why EigenLayer
QUEST needs cryptoeconomic security to be trustworthy as systemic infrastructure — a single operator is a single point of failure. EigenLayer’s restaking model is the only credible path to decentralized certification of a consensus-layer risk signal. The security model maps directly: operators stake on the correctness of Grey Zone Score calculations.
Contact: [email protected]
Public Repo:https://github.com/GuillermoSiaira/QUEST.git
Wallet: 0xBb3272F387dE5A2c2e3906d24EfaC460a7013f2C