DeFiScoreAudit detects insolvency, dishonesty, and fake growth in DeFi protocols early — grading solvency, legitimacy, and adoption on the same AAA-to-D scale institutions already trust, built on public on-chain data instead of quarterly filings.
Public ratings are free to read. No spam — new ratings, rating actions and methodology updates only.
You're on the list. New ratings and rating actions will land in your inbox.
New: chain ratings for Ethereum, Solana, TRON and six more networks, with how each network's revenue reaches its coin. Or read the full Aave rating with per-metric sourcing and four stress scenarios.
Traditional finance runs on a shared risk vocabulary: a BBB bond means the same thing to a pension fund in Tokyo and a bank in Toronto. DeFi asks every participant to be their own analyst — and the losses show it.
DeFi's biggest losses split into three failure modes — insolvency, dishonesty, and irrelevance. Blending them into one number is how risk gets hidden. We publish all three, separately, with the reasoning behind each. The Solvency grade is the credit grade; Legitimacy sits beside it, and Growth is shown as a separate outlook; neither is averaged into it.
Will it stay solvent and honor its obligations?
The bond scale the world already speaks, rebuilt for on-chain reality: collateral quality, contract risk, treasury runway, governance control, and liquidity structure. The BBB line carries the same weight it does in fixed income — investment grade versus speculative.
Is it actually what it claims to be?
Promise versus delivery: roadmap commitments checked against shipped code, treasury spending against stated use, disclosures against on-chain conduct. No collateral ratio catches a lie — this score exists because rug pulls are legitimacy failures, not solvency failures.
Is it growing toward its own stated goal?
Not vanity TVL — traction measured against each protocol's own mission. Organic borrow demand for a lender. Real settlement volume for a stablecoin. Liquidity that stays after incentives end. Graded on the same scale, published separately, never a buy signal.
Today every rating is a point-in-time snapshot, stamped with the date of its data. The surveillance layer below is in development; until it ships, grades change only when a dated re-screen is published.
Our purpose is to make protocol risk quantifiable, comparable, and visible before it materializes — so that capital allocators, protocol teams, and integrators in decentralized finance can act on measured risk rather than sentiment. DeFi removed intermediaries; it did not remove counterparty, collateral, or execution risk. We measure what remains.
Our mission is to be decentralized finance's reference source for early, quantitative risk assessment. Protocol risk today is assessed through fragmented signals — TVL charts, audit PDFs, governance forums, social sentiment. DeFiScoreAudit consolidates these into a single analytical framework: a factor-based quantitative model producing standardized credit grades, backed by published methodology. Continuous on-chain surveillance, with rating actions triggered by verifiable state changes at the protocol level, is the layer we are building next.
DeFi settles in blocks, but its risk assessment still runs on quarterly-report instincts. Collateralization ratios, liquidity depth, oracle dependencies, validator and treasury concentration — the data to price protocol survival exists on-chain, in real time, and largely goes unread. DeFiScoreAudit exists to read it systematically. Our end state is a rating architecture native to the settlement layer itself: grades derived from live chain state, updated on evidence rather than on a calendar, and consumable both by humans and by the systems built on top — wallets, aggregators, risk engines, underwriting models. When protocol risk is machine-readable infrastructure, failures stop being surprises and capital prices risk correctly.
AAA-to-D grades quantifying a protocol's capacity to remain solvent through stress. Derived from a weighted multi-factor model across collateral quality, liquidity, leverage, counterparty exposure, and contract risk — with notching caps that bound the grade wherever a single structural weakness (e.g., concentration or admin-key risk) dominates the profile.
A delivery-versus-declaration assessment: roadmap execution rate, audit disposition and remediation, team verifiability, and governance integrity, graded on the same AAA-to-D scale.
Usage measured against the protocol's own stated objective, filtered for incentive-driven volume: organic transaction share, retention cohorts, and integration growth, graded AAA to D against category peers.
Continuous monitoring of on-chain state against defined thresholds. Breach events — collateral ratio deterioration, abnormal outflows, concentration spikes, governance anomalies — trigger review and, where warranted, a published rating action with block-level evidence cited.
Deterioration is a process with observable precursors. Our thresholds are calibrated to flag divergence while intervention or exit is still possible — not to narrate failures after finality.
No engagement fee, grant, or token allocation can create, raise, or delay a grade. Independence is enforced by the revenue structure, not by policy language.
Every factor input resolves to verifiable on-chain state or documented, checkable fact. Narrative, roadmap promises, and market sentiment carry zero model weight.
A rating is a probabilistic opinion about survival under stress, not a guarantee against loss and not a return forecast. We publish what the model cannot see — and we keep our misses on the record, unedited.
The deliverable isn't the grade; it's the risk that gets fixed or avoided because of it. Each published rationale doubles as a remediation map for the protocol it grades.
Waitlist members get the first rating actions, the full published methodology, and early access to coverage requests.
New full ratings and rating actions only. Nothing else.
You're on the list. New ratings and rating actions will land in your inbox.
