CIBIL style trust for wallets
- One summary per address: 300–850, tier, driver lines.
- Like CIBIL style bureaus but from on-chain behaviour not loan paperwork.
- Used when teams judge treasury payouts, hires, protocol counterparties.
- Informal reputation and private lists scale poorly and hide why someone passed or failed.
- One chain history baseline helps honest wallets move faster and sends risky ones to review.
- One signal among many not identity judgement or repayment promise.
What we built
- UI for a bureau style trust read on an Ethereum address.
- Paste or connect wallet band tier driver bullets.
- Docs plus demo prove the loop.
How it works
- Valid EVM address wallet connect can autofill.
- Fetch score with address.
- Returns score tier highlights shown as bar badge list.
- Trust preview from behaviour not loan repayment.
How it’s made
- Next React Tailwind RainbowKit wagmi viem.
- Radix primitives Framer Motion.
- Metadata OG Twitter sitemap robots manifest.
Future work
- Backend indexes txs protocols recency risk weighted pillars.
- Thin file caps published methodology appeals.
- Anti gaming monitoring policy overrides.
- Consent plus integrations where score is one input.
Why Web3 wants something like this
Today, teams rely on allowlists, Discord reputations, snapshot votes, or gut feel before sending treasury funds, hiring contributors, or opening credit lines in DeFi. None of that scales. A portable wallet score—used as one input among many—could reduce friction for honest actors and surface wallets that look rushed, recycled, or entangled with toxic flows (exploits, launderers, sybil farms). It is not a promise that someone will repay a loan; it is a structured lens on history and hygiene.
What typically feeds such a score
A scorer turns public ledger facts into a few pillars, scores each pillar, then combines them into one band (e.g. 300–850) plus short reason lines. The home demo uses placeholder math; a production model would follow the same shape—just backed by real ingestion rules.
Nothing here is mandatory: pick pillars and weights to match your risk policy.
| Pillar | What we measure (on-chain) | Why it affects trust |
|---|---|---|
| Activity | How often the wallet sends txs; counts over 30 / 90 / 365 days; failed vs successful txs. | More history usually means statistical confidence. Very little activity isn't “evil”—it's a thin file (harder to judge). |
| Protocol breadth | Distinct contracts/apps touched—DEX, lending, staking, NFT venues, bridges—not just one loop. | Breadth suggests real participation; repetitive dust or wash loops can be discounted or penalized. |
| Recency | Last-seen time; long gaps; sudden bursts after silence. | Shows whether behaviour is continuous; odd patterns may trigger review instead of a tiny score tick. |
| Risk hygiene | Counterparties and flows—proximity to known exploits, launder patterns, policy lists you maintain. | Often hard negatives: they can outweigh many soft positives, like bureau exclusion logic. |
| Maturity | Wallet age, cumulative txs/nonce; optionally ENS or attestations. | Separates brand-new throwaways from accounts with a long, observable trail. |
| Optional ML | Clustering / anomaly flags over the full tx graph—still audited against rules. | Helps prioritize review; outputs should still map to explainable drivers, not a silent score. |
Each pillar becomes a sub-score; you choose weights, sum or blend, then cap outcomes when data is sparse so the headline number never overclaims—same intuition as thin-file handling in credit bureaus.
About this site
Wallet Score Checker is a small interactive front-end: you paste or connect a wallet and see a score-shaped response. Numbers are illustrative unless you plug in your own scoring service—the goal here is to communicate the idea of bureau-like clarity meeting Web3 openness, not to certify real-world credit.