Open sourceSelf-hostedExperimental

Protect. Observe. Respond.

Protective DNS with visibility into what your network is asking for.

Block known malicious domains at the resolver, investigate DNS activity, and understand the decisions behind your protection — without sending query telemetry to a SaaS platform.

See how it works ↓
● no account● no trial● no subscription
dnsdaddy · consoleProtected

DNS queries

24,918

last 24 hours

Threats blocked

312

feed + policy matches

Resolver status

Healthy

udp/53 · doh · dot

Query activity

live

A bar chart of DNS query volume over time, with blocked queries highlighted.

Recent DNS queries with client and verdict
TimeClientVerdict
14:02:11laptop-07allowed
14:02:09phone-miablocked
14:02:04nas-01allowed
14:01:58iot-cam-3blocked
14:01:52laptop-07allowed
Illustrative representation of the DNS Daddy console. Figures shown are examples, not benchmarks.

How it works

One control point. Every name your network resolves.

DNS Daddy sits between your devices and the resolver they would otherwise use. Nothing is installed on the endpoint — if a device uses your DNS, it is covered.

  1. 01

    Device

    A laptop, phone, server or IoT device asks for a name.

    →
  2. 02

    DNS Daddy

    Your self-hosted resolver receives the query over DNS, DoH or DoT.

    →
  3. 03

    Protect

    Policy and documented threat feeds decide whether the name is allowed or blocked.

    →
  4. 04

    Observe

    The outcome is logged locally with the client and a reason you can inspect.

    →
  5. 05

    Respond

    Use the evidence to investigate a query, change policy or review an allow-list decision.

PROTECT

Threat-feed matches and policy decisions perform blocking.

OBSERVE

Query activity and client context remain on infrastructure you control.

RESPOND

Evidence gives you a starting point for investigation and policy changes.

Explore the architecture →

Why DNS Daddy

Protective DNS without the black box.

Public protective resolvers can block known threats. DNS Daddy is for operators who also want to see which device asked, understand the decision, and keep the record local.

Public resolver

  • Easy to get started
  • Limited local visibility
  • Provider operates telemetry

DNS Daddy

  • Self-hosted
  • Per-network visibility
  • Local telemetry
  • Inspectable policies
  • Open source

Commercial protective DNS

  • Mature platforms
  • Larger feature sets
  • Managed service / SaaS model

DNS Daddy does not claim equivalent maturity, scale or support to established commercial platforms.

Threat intelligence

Known threats, blocked at the resolver.

DNS Daddy can ingest public and community-maintained threat intelligence for categories such as malware, phishing, command and control, and cryptomining.

Threat intelligence

Public feeds

active
community-malware24,918active
phishing-army18,402active
local-blocklist1,284custom

No black-box intelligence. Every built-in source is documented and inspectable, and you can add local feeds in supported formats.

Explore threat intelligence →

Behavioural detection

Six experimental detectors that observe, explain and alert.

DNS Daddy includes six experimental behavioural detectors designed to surface suspicious DNS patterns. They observe, explain and alert — they do not automatically block traffic. Enforcement stays with policy and curated threat feeds.

dns_tunnel

max high

DNS tunnelling

DNS used as a data carrier

dga_like

max high

DGA-like domains

Algorithmically generated names

nxdomain_anomaly

max medium

NXDOMAIN anomalies

Bursts of failed lookups

txt_anomaly

max medium

TXT anomalies

Unusual TXT record usage

dns_beaconing

max medium

DNS beaconing

Fixed-cadence check-ins

resolution_failure

max medium

Resolution failures

Domains persistently failing upstream

Experimental · alert-only

Thresholds are currently calibrated against synthetic test corpora rather than large-scale production traffic, and no measured real-world false-positive rate has been published yet. That is precisely why detections remain alert-only.

There is no machine learning here, and the project deliberately avoids relabelling statistics as AI. A finding shows the measurements behind its score, so you can judge it rather than trust it.

Every detector reports its own maturity through /api/v1/detectors, so the running build states this rather than relying on a web page being current.

External intelligence

Bring your own threat intelligence.

Alongside the curated blocklist feeds, DNS Daddy can attach external reputation and enrichment services — VirusTotal, Google Safe Browsing, or any custom HTTP/JSON provider. This is optional and disabled by default; an operator chooses whether a provider participates.

Credentials encrypted at rest

AES-256-GCM sealing, handled by a dedicated keyring.

Write-only credentials

Secrets are not returned by the management API after configuration.

Connection testing

Check a provider answers before it participates in anything.

Health monitoring

Per-provider status, so a silent failure is visible.

Rate limiting and circuit breaking

A struggling provider is backed off, not hammered.

Caching and bounded timeouts

Two-layer cache; a lookup never waits beyond its budget.

Reputation modes

  • off

    External services are not consulted at all.

  • cache_only

    Only previously cached reputation data is used.

  • blocking

    A provider may be briefly consulted during a DNS decision, within a bounded timeout.

Configuration establishes the maximum permitted mode. The dashboard can lower the operating mode during use; it does not raise it past that ceiling. A provider failure or timeout is treated as unknown, never as malicious.

What enabling a provider costs you

When an external provider is actively queried, the domain being investigated may be sent to that third party — the actual name, at the time it was asked for.

Nothing is sent anywhere unless you turn a provider on. This behaviour is opt-in, and the trade-off is documented in full rather than buried.

Read docs/external-apis.md →

Evidence & decisions

Why was this domain blocked?

Blocklists, behavioural findings and external reputation providers used to describe evidence in incompatible ways. DNS Daddy now has a shared evidence model, so a claim from any of them can be read the same way.

Evidence model

  • observationWhat was actually seen.
  • sourceWhich named source reported it.
  • timestampWhen it was observed.
  • claimWhat that source asserts.
  • confidenceHow strongly it asserts it.
  • expiryWhen the claim stops counting.
  • detailThe supporting specifics behind the claim.

Corroboration is expressed as agreement between identifiable independent sources — not as a single combined risk score with the working hidden inside it.

Decision records

DNS eventevidencedecisionexplanation

DNS Daddy can persist the reasoning behind an enforced DNS decision: which rule fired, the policy path it took, and the evidence cited at that moment. The explanation is written when the decision happens rather than reconstructed later, so a feed that changes tomorrow cannot quietly rewrite why something was blocked today.

Read docs/decision-records.md →

Not on by default

Decision recording is off unless an operator enables it, and it writes a record per blocked query rather than per query. Do not assume an existing deployment is already keeping these records.

Research

Daddybound — experimental DNSSEC validation research.

Daddybound is a DNSSEC validation engine being written from the DNS and DNSSEC standards in pure Go, and tested against established validators in a differential laboratory.

Not connected to the resolver

Daddybound answers no user DNS queries, enforces no DNS policy, and does not provide local DNSSEC validation to deployed DNS Daddy resolvers. A test in the repository asserts that nothing on the query path can even reach it.

DNS Daddy still relies on its upstream resolver for operational DNSSEC validation. Daddybound is an isolated experimental validation engine being developed and tested separately from the live resolution path.

Where the work has got to

  • Positive chain-of-trust walk from a configured trust anchor to a signed answer.
  • A differential laboratory comparing results against established validators.
  • No NSEC or NSEC3, so no authenticated NXDOMAIN or NODATA yet.
  • It is not, and must not be read as, production DNSSEC validation.

A separate point worth keeping straight: DNS Daddy can record the DNSSEC status reported by a validating upstream. That is a strictly weaker statement than validating signatures locally, and it is not presented as the same thing anywhere on this site.

Assurance

Evidence, not assumptions.

DNS Daddy is AI-assisted and says so plainly. Generated code is not automatically trusted code; claims are scoped to the evidence available today.

  • Automated test suite

    Re-checked automatically in CI.

    VERIFIED
  • Race detector

    Re-checked automatically in CI.

    VERIFIED
  • Static analysis

    Re-checked automatically in CI.

    VERIFIED
  • Vulnerability scanning

    Exercised by tooling at a known point in time.

    TESTED
  • Behavioural detection

    Working concept; insufficient real-world validation.

    EXPERIMENTAL
  • Independent professional security review

    No evidence currently supports this claim.

    NOT VERIFIED
NOT VERIFIED

There has been no independent professional security review.

This project is experimental software, not audited or production-grade security software.

Read the assurance evidence →

Live threat intelligence

See DNS-related threats emerge, connect and evolve.

Explore DNS-related malicious infrastructure using real threat-intelligence observations, infrastructure relationships and defensive context — through the DNS Daddy Threat Observatory.

Real observations · Infrastructure relationships · ATT&CK context · Defensive guidance

Open Threat Observatory →

threats.dnsdaddy.dev

An experimental DNS threat-intelligence and visualisation project built as part of the wider DNS Daddy security research ecosystem. This page shows only the Observatory's own sanitised public output — no provider feeds or credentials live here.

Live threat activity

Observatory

DNS Daddy Threat Observatory

Explore live DNS-related threat intelligence, infrastructure relationships and defensive context.

Open Threat Observatory →

The DNS Daddy ecosystem

Understand DNS. Protect DNS.

Protect

DNS Daddy

Open-source protective DNS experimentation you run yourself.

Observe

Threat Observatory

Explore real DNS-related threat intelligence and the infrastructure behind it.

Explore live data →

Learn

Research & methodology

Understand how DNS threats are observed, correlated and defended against.

Self-hosted deployment

Run it where you want.

Designed to run on modest hardware, with deployment testing still ongoing. Choose an environment you understand and can maintain.

01

Docker

Containerised deployment

02

LAN / homelab

Local network resolver

03

Public VPS

SSH tunnel or HTTPS

04

Your infrastructure

Local telemetry and control

install-dnsdaddy
$ git clone https://github.com/jameshoulder/dnsdaddy.git
$ cd dnsdaddy
$ ./deploy/install-docker.sh

Works alongside Pi-hole

Already using Pi-hole? Keep it.

Pi-hole

Ads and trackers

Pi-hole is an excellent choice for network-wide ad and tracker blocking.

DNS Daddy

Threat intelligence and visibility

DNS Daddy focuses on protective DNS, documented threat feeds and DNS security visibility. They can run together.

DNS Daddy + Pi-hole guide →

Open source / built in public

A well-described bug is more useful than pretending the project is finished.

DNS Daddy is open because security software benefits from inspection, challenge and contribution. Read the source, review the decisions, and help improve the gaps.

Apache 2.0 licensed
Public source code
Public threat model
Documented security testing
Public roadmap
Contributions welcome

Source of truth

Everything shipped on this site should be traceable to the repository or clearly marked experimental, planned or not verified.

View DNS Daddy on GitHub ↗

Free & Open Source · Self-Hosted · No Account · No Subscription

Limitations

What DNS Daddy does not claim.

No independent professional security review
Behavioural detections are experimental and alert-only — they never block
No local DNSSEC validation; DNS Daddy records a validating upstream's verdict
Daddybound is isolated research and is not wired to the resolver
No native SIEM connectors — structured findings are exported as NDJSON for a log shipper
No machine learning anywhere in the project today
It cannot stop a client that resolves through an external DoH or DoT service
It cannot see cloud firewall or security-group state, and never claims one
No clustering, anycast, SSO, RBAC or multi-tenancy
safeSearch is accepted by the API but not enforced by the resolver

Protect. Observe. Respond.

See what your network is asking for.

Open source. Self-hosted. Built in public.