The reliability layer for AI agents

Hope Is Not a Retry Strategy.

Quorvel gives your AI agents a durable memory, exactly-once execution, and human approvals — so a crash, a duplicate, or a bad decision never becomes a disaster.

$npm i @quorvel/core
Star on GitHub ● Open SDK · MIT Building in public · v0.1
No card required 3-line integration Works with any LLM stack
quorvel — live action ledger streaming

Built to work with your stack — integrations, not endorsements

ledger
charge_card committed
send_email deduped
approval
issue_refund $420
✓ approved by blue
alert
payment.failed → retried ✓
recovery
crash → resumed
exactly-once · 0 dupes
One system, every surface

Every action, on one durable stage.

Ledger, approvals, recovery, and alerts — one source of truth, orbiting wherever your agents act.

How it works

Three primitives. Never start over.

Wrap your agent’s actions once. Spin the dial to move through recording, gating, and recovery — the page keeps scrolling normally.

spin · drag ↕
action
inputs
status
idem_key
idem_9f2a3c… · written before any side effect fires
01 · RECORD

Persist before you run

Every action is written to a durable ledger with its inputs, outputs, status, and a cryptographic idempotency key — before a single side effect fires.

awaiting approval
✓ approved — shipping
02 · GATE

Pause what’s risky

Policies decide what needs a human. Risky actions wait for approval from Slack, the dashboard, or your own UI — approve, reject, or edit before it ships.

03 · RECOVER

Resume or roll back

Crash mid-flow? Resume exactly where you left off. Step failed? Compensations for prior steps run in reverse — making failure recoverable.

The difference between a demo and production is what happens when something breaks.

Features

Make failure recoverable.

Six guarantees that turn a flaky agent into one you can ship to production.

action
inputs
status
idem_key

Durable Ledger

Every action persisted before it runs — inputs, outputs, status, and a cryptographic idempotency key. Your agent’s black box recorder.

3 attempts → 1 commit

Exactly-Once

Retries, duplicate webhooks, and double-clicks collapse into a single committed action. No more accidental double refunds.

issue_refund$420.00
ApproveReject
✓ Approved by blue

Human-in-the-Loop

Set a policy and risky actions wait for approval from Slack, the dashboard, or your own UI. Approve, reject, or edit before it ships.

1234
step 4 failed — rolling back 3 · 2 · 1

Auto-Rollback Sagas

When step 4 fails, Quorvel runs your compensations for steps 1–3 in reverse. Partial workflows clean up after themselves.

⚡ CRASH
resumed from checkpoint · exactly-once

Crash-Safe Workflows

A process dies mid-run? Quorvel replays from the last committed checkpoint — no lost work, no double-spends, no manual cleanup.

⚠ payment.failed → auto-retried ✓

Real-Time Alerts

Failures, approvals, and stuck workflows ping you in Slack or email the instant they happen — with a deep link straight to the action.

Mission Control

See every action as it happens.

One dashboard for the whole lifecycle — live ledger, pending approvals, replays, and alerts. No more guessing what your agent did.

LedgerApprovalsAlerts
live
Live ledger
Approvals 3
Sagas
Replays
Alerts
API keys
org: acme · pro
Actions today
12,840
Exactly-once
100%
Awaiting approval
3
Recovered
27
TimeActionIdempotencyStatus
14:22:07issue_refund · $420idem_9f2a…approved
14:22:01charge_card · $89idem_7c10…committed
14:21:58send_email · receiptidem_7c10…deduped
14:21:50sync_crm · contactidem_3b8e…committed
14:21:42export_invoiceidem_1a55…running
Runs anywhere

Reliable everywhere your agents run.

Cloud, edge, or your own VPC — every action flows through one durable ledger. Designed so replays and approvals stay consistent no matter where the work runs.

multi-regionedgeyour VPCself-hosted
For developers

Three lines to production.

Wrap any side effect in run(). Quorvel records it, dedupes it, gates it, and recovers it — you keep writing plain TypeScript. Same primitive, every stack.

Vercel AI SDKOpenAILangChainMCPTypeScriptNode
import { HostedLedger, run } from '@quorvel/core'

const ledger = new HostedLedger({ apiKey: process.env.QUORVEL_KEY })

// wrap a Vercel AI SDK tool — durable + exactly-once
const refund = tool({
  parameters: z.object({ charge: z.string(), amount: z.number() }),
  execute: (args) => run(ledger, 'issue_refund', () =>
    stripe.refunds.create(args))
})
Demo vs. Production

The gap nobody warns you about.

🤞 Hope as a strategy

  • A crash mid-workflow leaves half-finished state
  • A retried webhook charges the customer twice
  • The agent refunds $40,000 with nobody watching
  • “What did the agent actually do?” — no one knows
  • You find out it broke from an angry customer

✓ Quorvel in production

  • Replays from the last checkpoint — zero lost work
  • Idempotency keys collapse dupes into one commit
  • Refunds over $100 wait for a human approval
  • Every action is in the ledger, fully auditable
  • Alerts hit Slack the instant something fails
Why reliability, not hope

Most agent demos work because nothing goes wrong on stage. Production is the part nobody films — the retry that double-charges, the crash mid-refund, the action no one can explain afterward. Quorvel exists for that part: record before you act, gate what’s risky, recover what breaks. Reliability isn’t a feature you bolt on. It’s the difference between a demo and a product.

Pricing

Usage-based. Unlimited seats.

Start free, scale when you ship. Usage-based, never per seat — with a flat-rate Scale tier when you go big.

Free
$0 /mo

For prototypes and side projects.

  • 1,000 actions / mo
  • Durable ledger + exactly-once
  • 7-day history
  • Community support
Start free
Recommended
Pro
$29 /mo

For teams shipping agents to real users.

  • 100,000 actions / mo
  • Human-in-the-loop approvals
  • Slack + email alerts
  • 90-day history
  • Overage $5 / 50k
Start Pro trial
Scale
$99 /mo

For high-volume, mission-critical agents.

  • Unlimited actions
  • Auto-rollback sagas
  • Multi-region durable replays
  • 1-year history
  • Priority support
Choose Scale
Enterprise
Custom

For regulated and large-scale orgs.

  • SSO / SAML + audit logs
  • On-prem / VPC deploy
  • Custom retention
  • Dedicated support
Contact sales
Built to finish

Reliable AI starts here.

Stop hoping your agent makes it to the finish line. Give it a memory, a safety net, and a human in the loop — in three lines of code.