We manage your security to weather the AI storm →

Web3 MDR: managed detection and response

Guardian’s Web3 MDR is managed detection and response for crypto teams: engineers on your contracts, multisigs, cloud and domains around the clock. Each alert is triaged by someone who knows your system, with a 15-minute SLA on Critical alerts. It is the response side of Managed Security.

15 min SLA on Critical alerts
24/7 An engineer on every alert
35 min Median first public alert, 30 exploits
88 min Median pause, 19 of the 30
7 of 11 Drains over before any public alert

Trusted across Guardian engagements

  • GMX
  • USDT0
  • LayerZero
  • Ethena
  • Synthetix
  • Jupiter
  • Yuga Labs
  • Orderly Network
  • Veranta

Managed Security · Web3 MDR

What Guardian watches, onchain and off.

  • Contracts, every block Privileged calls, role and ownership changes, upgrades, and large fund movements.
  • Multisigs and signers Multisig transactions, each verified before it is signed, against a threat model of every path to your funds.
  • Detections built for you Monitors customized to your protocol, beside out-of-the-box detectors for suspicious contracts and known exploit patterns.
  • Deployments and configuration Every deployment and configuration change, reviewed as it lands.
  • Domains, DNS and certificates Watched by Sentry, so a changed record or certificate reaches an engineer.
  • Cloud and code host Your cloud accounts and code host, watched by Sentry beside the chain.
  • Internet-facing surface Shield runs reconnaissance and pentesting around the clock against the hosts, ports, services and storage you expose.
  • Security advisories Disclosures ingested in real time and matched against what you run.

Already run monitoring tools? Your existing monitoring streams into Guardian’s 24/7 MDR team, alongside Guardian’s own detectors.

How incidents are handled

From a detection to a response, not a notification

  1. 01

    Detect

    A monitor fires on chain state or an offchain change: a privileged call, an upgrade, a queued multisig transaction, a DNS record.

  2. 02

    Triage

    A Guardian engineer who knows your system picks it up, around the clock, under a 15-minute SLA on Critical alerts.

  3. 03

    Respond

    The engineer works the response with your team. Automated actions, such as a pause, run through authority you grant in advance.

  4. 04

    Patch

    Where a finding or advisory reaches your system, the engineer confirms the exposure and works the fix through with you.

Before any of it, onboarding locks things down: a complete map of your assets, a threat model of every path to your funds, Shield pentesting of everything facing the internet, and your single points of failure fixed.

Response authority

Who can act, agreed before an incident

No provider can act onchain by default. A pause needs a role, a pre-signed transaction or a module you grant and test first.

Read-only by default

Managed Security runs on least privilege. Monitoring and threat models run on public chain state, Shield sees only what an attacker sees, and Helix reads only the repositories you choose.

Read: how Managed Security works.

Automated actions

Guardian monitoring supports custom rules and actions, including automated pauses. Each runs only through authority you grant.

Read: emergency pause design.

Your team stays in it

Your team keeps full access to Guardian’s alerts, monitors and tools, and works each response with the engineer on it.

Read: writing a DeFi incident response plan.

Recent incidents

The minutes between an alert and a pause.

All three predate Managed Security. They show what the time to respond decides.

  1. $292MKelp DAO

    Poisoned RPC nodes fooled the only verifier on a Kelp route, releasing 116,500 rsETH. Kelp’s emergency multisig paused core contracts 46 minutes later.

    Apr 2026 Our breakdown of the Kelp DAO incident

  2. $25MResolv

    Resolv’s own monitoring flagged the first unbacked mint, but pausing took nearly three hours.

    Mar 2026 Our breakdown of the Resolv incident

  3. $3.8MNEAR Intents

    Detection came about 40 hours after the exploit began. The funds were later returned.

    Oct 2026 Our breakdown of the NEAR Intents incident

Published work

Signing, multisigs and response, on the record.

“Their keen eye for detail and responsive support are qualities that shine through in every engagement.”
Core TeamLayerZero

Questions

Web3 MDR questions

What is Web3 MDR?

A service, not a tool: a team watches your contracts, keys and infrastructure around the clock, triages every alert and responds. What is Web3 MDR? explains how it differs from traditional MDR and from monitoring.

What is Guardian’s response time?

A 15-minute SLA on Critical alerts, around the clock, onchain and offchain.

Can Guardian pause our contracts?

Only through authority you grant. Guardian monitoring supports automated actions, including pauses, and each needs a role or transaction set up with you first. Otherwise the service is read-only.

Do we have to replace our monitoring tools?

No. Your existing monitoring streams into Guardian’s 24/7 MDR team, alongside Guardian’s own detectors.

What does the 90-day guarantee cover?

Within the first 90 days, Guardian detects and resolves a high-severity issue with funds at risk or services that could be disrupted, or you don’t pay.

Which MDR providers protect crypto companies?

Our ranking of crypto MDR providers compares nine on response SLAs, authority to act onchain and offchain coverage. Guardian publishes it and lists itself first, disclosed.

Resource Hub

Before you choose an MDR provider.

Services

Other engagements

Start now

Get an MDR team that knows your protocol.

Get a quote