credit
you
own
a credit reputation you carry, not one a bank keeps. a k-of-n threshold oracle on stellar soroban turns your on-chain history into fair credit — portable, yours, signed by an independent quorum. zero fintech in the trust path.
live on-chain
reading…score any address, live
paste any stellar public key. the scoring engine reads horizon and returns the same features the oracle signs — no fintech, no private data.
watch the threshold sign
every mint requires k of n independent oracle signatures over the canonical 92-byte message. no single party can authorise alone.
env.crypto().ed25519_verify k times. fewer than k valid signatures, a duplicated oracle index, or any single tampered byte rejects the mint at simulation — no ledger pollution, no gas spent.architecture
three components, each independently verifiable. nothing in the trust path is owned by a single party.
synthetic scoring engine
off-chain reader of stellar horizon. 180-day window, 200-op cap. p2p churn discounted 70% against an ecosystem whitelist. zero fintech dependency.
threshold oracle quorum
five independent ed25519 keypairs. three signatures required. each signs the canonical 92-byte mint message. tampering one byte breaks all signatures.
soroban contracts
vigente-badge (soulbound credit token, threshold-verified mint, immutable slash) + reference-vault (credit-gated lending with TVL cap, util limit, withdrawal timelock).
threat model
six adversarial scenarios. each one has a code-level mitigation that ships in the current testnet contracts.
out-of-scope items (validator collapse, compromised user wallet, sdk bugs) are deliberately listed as such so the boundary of the protocol's responsibility is explicit.
partners we're building with
vigente is the credit primitive — the lending stack lives on top, the data layer feeds in. we're forming partnerships in two directions, with conservative limits and client protection baked in.
a protocol reads get_score / is_defaulted to open a reputation-tier pool. for price-oracle markets, vigente gates eligibility off-chain at conservative, throttled limits (CP2). reference-vault is the on-chain reference any protocol can read directly.
integrate vigente as credit layerthe core score reads only stellar horizon, so the trust path stays fintech-free. open finance is opt-in enrichment that adds detail to a thin-file borrower's profile — never a gate, never published on-chain.
partner on dataif you run a soroban lending market and want to plug vigente's credit primitive in front of your pool — same week integration, zero token swap.
inclusion is the product
a credit history is an asset the user owns — soulbound, portable, readable by every protocol that integrates vigente. it turns on-chain activity into fair credit. funded by the protocols that use it, not by grants — sustainable by design, not charity.
shared value, not charity
the underserved market — earners with on-chain or open-finance data but no traditional score — IS the target market. every score that unlocks fair credit earns a fee and cuts financial exclusion at the same time.
no over-indebtedness
conservative tier ceilings, a first-loan throttle to 10%, and immutable defaults. enabling credit to the unbanked must never push them into unpayable debt — the limits are proven on-chain in reference-vault.
metrics that aren't vanity
we measure first-time credit access, cost of credit before vs. after, real default rate, and 6/12-month persistence — with a baseline and external verification. not 'wallets created'.
fair, explainable, private
disparate-impact audits across groups, an explainable score with a right to human review (Ley 21.719), and no personal data published on-chain. client protection (Cerise+SPTF) is a release gate, not a slogan.
aligned to SDG 1 / 5 / 8 / 10 and the Cerise+SPTF client-protection standards. measured with IRIS+ — baseline, comparison group, external verification.
see it live on testnet
two real soroban calls against CDLLO7QE…. the negative one shows the age floor enforcing on-chain through the signed account_age bytes.
positive mint
tx on ledgerage-floor trap
rejectedroadmap
built in the open, funded in tranches. everything marked shipped is verifiable on-chain today — the rest is scoped, costed, and labeled by the tranche that pays for it.
the wedge is layer 1: a live lending protocol reading the score in production. distribution and the data layer compound from there.
credit primitive
- ·3-of-5 threshold oracle verified on-chain
- ·soulbound credit badge + immutable defaults
- ·credit-gated reference vault (TVL cap, timelock)
- ·credit oracle interface v1 + ABI for integrators
- ·180-day on-chain credit heat map
first protocol integration
- ·first lending-protocol integration (off-chain gate, tested)
- ·live reputation-tier pool at conservative limits
- ·oracle ops + key rotation runbook
- ·SEP draft: credit attestation standard
data + capital efficiency
- ·open finance enrichment (consented bank data)
- ·SEP-0056 tokenized vault + tokenized-vault listing
- ·idle reserve earning yield in soroban pools
- ·/earn — one-click USDC deposits for LPs
open infrastructure
- ·mainnet deploy behind multi-sig
- ·typescript SDK on npm
- ·tier-segmented pools + staking
- ·inclusion pilot + client-protection metrics (IRIS+)
deliberately out of scope until mainnet: own token, multi-chain, retail KYC, competing with existing lending markets. vigente is the credit layer other protocols read — not another lending app.