Connect agents.
Control execution.
Eraivo gives builders a governed way to add wallets, payments, escrow, settlement, approvals, and audit trails to agent and institutional workflows — through MCP tools or a TypeScript SDK.
// agent-side MCP tool call
{
"tool": "eraivo_pay",
"arguments": {
"to": "data-provider.eth",
"amount": "120.00",
"asset": "USDC",
"purpose": "market-data-subscription"
}
}
// governed result
→ decision: allow
→ policy: within-mandate
→ audit: recorded, hash-chainedFrom sandbox workflow to live controlled execution.
Choose The Workflow
Start with the action your product needs: agent payment, treasury movement, escrow creation, settlement, or approval-controlled wallet access.
Set The Rules
Define who can act, how much they can move, which counterparties are allowed, and when human approval is required.
Run In Sandbox
Test the full decision path before live execution: request, policy check, approval, hold, rejection, execution, and audit record.
Go Live
Connect the workflow to production once the controls, evidence, approval path, and audit output match your operating requirements.
One platform, six ways in.
MCP server
The primary agent interface: typed eraivo_* tools with structured errors, covering accounts, payments, escrow, approvals, commerce, and audit.
REST clients
EraivoClient, AgentAccountClient, and SettlementEngineClient for services that integrate directly over HTTP.
Direct-contract mode
Read registries, verify proofs, and build calldata against the protocol contracts without any Eraivo server in the path.
Relayer pool
Multi-relayer execution with failover, health checks, deduplication, and operator discovery.
Decentralized storage
Encrypt, content-hash, publish to IPFS, and verify evidence and audit bundles.
Policy hashing
Deterministic policy hashes so a mandate can be committed on-chain and re-verified by anyone.
Product-level capabilities, not low-level plumbing.
Let agents hold and move funds inside explicit permissions.
Operate wallet fleets with roles, limits, approvals, and reporting.
Hold funds against milestones, evidence, and dispute outcomes.
Coordinate payment and asset settlement with checks before execution.
Route sensitive actions to humans before money moves.
Require counterparties, purpose, evidence, and policy validation.
A record for every request, decision, approval, rejection, and execution — hash-chained and anchorable on-chain.
Policy commitments, settlement proofs, and audit anchors verifiable against on-chain registries.