ERC-8004 · Live on Base mainnet
Vaultfire × Helixa — the trust stack
Helixa identifies the agent. Vaultfire commits the agent. Same ERC-8004 base, complementary primitives, working code. Every number on this page comes from a live read against Base mainnet — no caching of state, no mocks.
Accountability bonds
—
0.0000 ETH escrow
Partnership bonds
—
0.0000 ETH escrow
Insurance pool
0.0000
ETH safety net
Reading live state from Base mainnet…
Walking Helixa identity, both bond contracts, and live escrow. Cold reads take 25–60s.
How this combines
Helixa — the identity layer
AnswersWho is this agent?
PrimitiveERC-8004 NFT + 11-factor Cred Score (0–100)
RecourseSoft — Cred drops, Soul fragments
Money$CRED token, x402 paid endpoints
Vaultfire — the accountability layer
AnswersWhat are they committed to?
PrimitiveAccountability + Partnership Bonds, Flourishing Oracle
RecourseCommitment decay + Dispute Resolution + Insurance Pool
MoneyETH-staked bond escrow, 71 paid x402 endpoints
Cred scores who they are. Bonds prove what they’ll commit. Both run on ERC-8004. Together they’re the trust signal agent commerce needs.
No slashing. Bonds are commitment stake feeding the Flourishing Metric Oracle and arbitrated through VaultfireDisputeResolution. Bonds decay; they aren’t punitively slashed. Positive-sum, on purpose.
One call. Four fields. No SDK.
Add &simple=1 to any trust query and you get a Stripe-shaped response: the answer to “is this agent trustworthy enough for this action?” in one HTTP call.
REQUEST
GET https://theloopbreaker.com/api/helixa/trust?address=0xA054f831B562e729F8D268291EBde1B2EDcFb84F&simple=1
RESPONSE
{
"address": "0xA054f831B562e729F8D268291EBde1B2EDcFb84F",
"combinedScore": 43,
"badge": "committed",
"hasIdentity": false,
"bondCount": 6,
"stakedETH": "0.002400",
"flourishingScore": 59,
"enforcement": "commitment + dispute resolution, not slashing"
}Try it live →·No auth. No SDK. CORS-open.·~200ms p50 from Base mainnet. What Helixa gets — phased integration
Three phases, each independently valuable. Phase 1 ships the day we agree — no token gating, no fee, no data hand-off. Each phase compounds the next.
PHASE 01LIVE TODAY
Bond posture in Helixa profiles
Ships day-one
One HTTP call surfaces Vaultfire bond data inside any Helixa agent profile or the Agent Terminal — already wired and serving traffic on this page.
HELIXA GETS
New trust column for free — “Cred 80 + 0.0024 ETH at stake”
Prime tier becomes harder to fake — bonds gate the top score
$CRED token gets a reason to exist beyond reputation
PHASE 02CODE READY
Bonded handshakes
A few hours of frontend on your side
Helixa's HandshakeRegistry entries get a one-click escrow button that calls AIPartnershipBondsV2. Adapter is compiled and tested.
HELIXA GETS
Higher-value handshakes — counterparties trust escrow
New revenue surface — take a thin fee on bonded handshakes if you want
Network effect — every bonded handshake makes the next one safer
PHASE 03DESIGN AGREED
Joint trust score endpoint
Co-designed with you
A single combined Cred × Vaultfire score that partners actually want to integrate. Helixa-branded, Helixa-controlled, served from your domain.
HELIXA GETS
You own the canonical agent trust score in the ecosystem
Partner protocols (CDP, AgentKit, A2A) plug into one number, not two
You set the weights — we’re infrastructure, not a competing brand
The ask is small and the value compounds. Phase 1 is a one-line API call — zero risk, zero deploy, zero token gating, no fee, no data hand-off. Vaultfire stays the accountability layer underneath Cred. Helixa stays the identity brand on top. Each phase you ship makes both stacks stronger.