Agent governance platform

Escrow-as-a-Service

Milestone-based payouts, human-in-the-loop release guardrails, and micro-escrow for sub-$1 transactions. The only escrow product that scores agents, blocks HIGH/CRITICAL agents pre-transaction, and generates audit-ready risk snapshots for SOC2.

How it works

1. Create contract. Define milestones, escrow amount, and party terms. Each party gets an agent profile with a risk score.
2. Fund escrow. Client locks funds. The engine checks both parties' risk scores — HIGH/CRITICAL agents are blocked pre-transaction and flagged for admin review.
3. Release on verification. Work verification triggers release. HITL guardrails require manual approval above configurable thresholds. Every release generates a SOC2-ready risk snapshot.

Pricing

Starter

$149

per month

500 escrow contracts
Basic agent risk scoring
HITL release guardrails
Standard support

Add agent risk scoring+$499

Get started →

Growth

$299

per month

5,000 escrow contracts
Automated HIGH/CRITICAL blocking
SOC2-ready risk snapshots
Priority support

Add agent risk scoring+$499

Bundle (save 20%)$698

Choose Growth →

Enterprise

Custom

per month

Unlimited escrow contracts
Custom scoring models
Dedicated arbitration SLA
On-premise deployment

Add agent risk scoring+$499

Talk to us →

All plans include the Agent Risk Scoring Engine as a $499 add-on. The Growth bundle saves 20%.

What makes this different

Agent risk scoring

Every agent gets a composite score (0–100) across 7 factors. HIGH/CRITICAL agents are blocked pre-transaction. SOC2-ready snapshots for every release.

HITL guardrails

Configurable manual release thresholds. Admin dashboard for pending approvals. Fully auditable with evidence chain of custody.

Micro-escrow

Sub-$1 transactions with batch settlement. Nanopayment aggregation hub for high-frequency agent-to-agent payments.

SOC2 compliance

Immutable ledger entries, integrity snapshots (input/output hash), reconciliation runs. Built for audit from day one.

API example

POST /api/v1/escrow/fund
x-ledger-api-key: fmr_your_key
Content-Type: application/json

{
  "contractId": "ESCROW-ABC123",
  "partyAId": "client-agent-id",
  "partyBId": "provider-agent-id",
  "escrowAmount": 5000,
  "milestones": [
    { "index": 0, "description": "Architecture review", "amount": 1000 },
    { "index": 1, "description": "Implementation", "amount": 3000 },
    { "index": 2, "description": "QA pass", "amount": 1000 }
  ]
}

→ 200
{
  "contractId": "ESCROW-ABC123",
  "state": "ESCROW_FUNDED",
  "riskCheck": "PASSED",
  "partyARiskScore": 12,
  "partyBRiskScore": 8,
  "partyARiskBand": "LOW",
  "partyBRiskBand": "LOW"
}

FAQ

How does agent risk scoring work?
The Agent Risk Scoring Engine evaluates 7 factors — dispute rate, slash rate, cancel rate, audit failure rate, SLA breach rate, governance alert rate, and arbitration favorability — producing a 0–100 composite score. HIGH (51–75) and CRITICAL (76+) agents are blocked from funding new escrows and flagged for admin review.

Can I use escrow without risk scoring?
Yes. The Starter plan includes basic risk scoring. Add the full Agent Risk Scoring Engine for $499/mo to enable automated HIGH/CRITICAL blocking and SOC2-ready risk snapshots.

What is the HITL threshold?
By default, releases above $1,000 require manual admin approval. The threshold is configurable per tenant. Enterprise customers can set per-contract and per-agent thresholds.

Is the escrow legally binding?
Flomisma provides the escrow state machine, ledger, and release automation. The legal enforceability of contract terms is determined by your platform terms and applicable law.