Skip to content
NewCybercrime Report 2026 · Eight Chapters · A Post Is Not An Incident → Explore The ChaptersCybercrime Report 2026 → Explore

Odin Research / What is / Offensive and testing tools

What Is Interactsh? The OAST Callback Server Explained

The callback server testers use to confirm blind bugs, and attackers use to confirm exploits. How it works and what to watch for in your egress logs.

By
Odin by Brandefense
Updated
21 Nov 2026
Reading time
8 min read
Type
OAST service (dual-use)

In short

Interactsh is an open-source out-of-band application security testing (OAST) tool from ProjectDiscovery. It generates unique domains and logs any DNS, HTTP, SMTP, LDAP or FTP request made to them. Testers use it to confirm blind vulnerabilities such as SSRF. Attackers use the same callbacks to learn which exploit attempts reached a vulnerable server.

Key takeaways

What it is
An MIT-licensed OAST server and client from ProjectDiscovery that records out-of-band interactions such as DNS lookups and HTTP requests.
How it works
A probe carries a unique Interactsh domain. If the target server resolves or fetches it, the Interactsh server logs the callback and the tester knows the bug is real.
Why it matters
Unit 42 saw attackers abuse Interactsh two days after its public release in 2021. Callbacks from your servers can mean someone is probing them.
What Odin Research saw
1,245 Interactsh endpoints in Monograph 17, 638 of them confirmed or suspected nodes. A few share hosts with C2 frameworks; most are ordinary testing servers.
What to do first
Stop internal servers from making direct outbound DNS and HTTP connections, and alert on lookups of OAST domains from server subnets.

What is Interactsh?

Interactsh is an open-source tool from ProjectDiscovery that detects out-of-band interactions. It generates unique domains and logs every DNS, HTTP, SMTP, LDAP or FTP request made to them. Testers use it to confirm blind bugs. Attackers use it to confirm which exploit attempts landed.

The technique is called out-of-band application security testing, or OAST. PortSwigger defines OAST as dynamic testing that uses an external server the tester controls to expose flaws ordinary scanning cannot see. Typical examples are blind SQL injection, blind cross-site scripting, blind command injection and server-side request forgery (SSRF).

The project's README says Interactsh was inspired by Burp Collaborator. It is written in Go and released under the MIT license. It is one of the most widely used OAST tools; see our overview of OAST for the wider category.

Interactsh is dual-use. A running Interactsh server most often belongs to a tester or a bug-bounty researcher. Seeing one is not a sign of attack. Seeing your own server call one is different.

How Interactsh works

Interactsh has three parts, according to its README:

  1. interactsh-client. A command-line tool and Go library. It registers with a server, generates unique payload domains and polls for interactions.
  2. interactsh-server. The listening side. It runs DNS, HTTP(S), SMTP(S), LDAP and FTP(S) services under one wildcard domain. Self-hosted servers can also run SMB and capture NTLM hashes Responder-style.
  3. Web client. A browser dashboard that shows interactions as they arrive.

A test runs in four steps:

  1. The client asks the server for a unique subdomain.
  2. The tester plants that subdomain in a probe, for example in a URL parameter, an HTTP header or an XML entity.
  3. If the target is vulnerable, the back end resolves or fetches the subdomain. That request reaches the Interactsh server.
  4. The client polls the server, sees the interaction and links it to the probe.

ProjectDiscovery runs default public servers, which its README warns can change or go down. Anyone can also self-host a server with a domain and a small VPS. The Nuclei scanner has built-in OAST support through Interactsh, which is why so many template scans trigger these callbacks.

How attackers use Interactsh

Attackers face the same problem as testers. A blind exploit gives no visible answer. A callback tells them the exploit worked, without dropping any malware first.

  • Abuse started almost at once. Palo Alto Networks Unit 42 reported in October 2021 that Interactsh became public on 16 April 2021. Unit 42 saw the first abuse on 18 April 2021.
  • Volume. Unit 42 counted about 32,200 Interactsh hits in its URL filtering data between March and September 2021. Many came from groups of addresses using the same URL, which looked like scanning rather than research.
  • Old bugs still pay. The most targeted flaws in Unit 42's data included CVE-2017-9506, an SSRF flaw in the Atlassian OAuth plugin, and CVE-2017-12629, a remote code execution flaw in Apache Solr.
  • Private OAST hosts. In November 2025, VulnCheck described about 1,400 exploit attempts across more than 200 CVEs from one private OAST host on Google Cloud. Most attempts looked like standard Nuclei templates.

Unit 42 summed up the defender's view well: each unique Interactsh URL can act like a C2 address, because it tells the operator which target answered.

What Odin Research observed

Interactsh is one of the most common tools in Odin Research's data. The key finding is that almost all of it is ordinary testing infrastructure.

  • A large, mostly benign fleet. Monograph 17 mapped 1,245 Interactsh endpoints. Of these, 638 are confirmed or suspected Interactsh nodes: 523 confirmed, 112 suspected and 3 on dynamic ISP lines. Most run on Amazon (360) and DigitalOcean (180), in the United States (575) and Germany (145).
  • A few bad neighbors. Some Interactsh listeners share an IP with other offensive tools. Published examples include hosts with Cobalt Strike, VShell, ARTEX and an ezXSS panel. In Monograph 07, an exposed Sliver workbench also ran a dual-use Interactsh server.
  • A major look-alike. Monograph 13 found 356 AWS Application Load Balancers labeled as Interactsh. 189 of them returned the exact TLS fingerprint seen on AWS-hosted Interactsh servers. The fingerprint reflects the AWS TLS stack, not the tenant.
  • Feed background. In Monograph 02, Interactsh accounted for 27 records in a scanner and OAST group of dual-use tools.

Monograph 13 sends the 523 real Interactsh servers to hunting rather than blocking, because the tool is dual-use. That is the right default for most teams.

How to detect Interactsh

Two different questions need two different detections.

Is my server calling an OAST host? This is the one that matters most.

  • Alert when web, middleware or database servers resolve random-looking labels, often 26 to 36 characters long, under known OAST domains.
  • Watch for outbound DNS, HTTP, SMTP or LDAP from server subnets that should never talk to the internet.
  • Correlate a callback with the request that preceded it. A web worker that fetches an OAST URL right after an odd inbound request is being probed.

Is this host an Interactsh server? Useful for hunting and for checking your own estate.

  • Interactsh servers answer with an X-Interactsh-Version header or Server: Interactsh-Server.
  • A real server reflects the correlation ID of a test request, exposes a /register API and listens on several ports at once, typically 53, 80, 443, 465 and 587.
  • Do not rely on JARM alone. On AWS, load balancers return the same value.

Odin Research publishes rules for both questions. Monograph 13 has Suricata signatures for the Interactsh banner and for /poll?id= polling. Monograph 17 adds Suricata signatures for Interactsh headers and client polling. It also has a Sigma rule for interactsh-server outside approved red-team hosts, a YARA rule for self-hosted server configs and a KQL hunt for server-tier OAST egress. All are in the Defense Kit.

How to defend against Interactsh-based probing

  1. Cut server egress. Block direct outbound DNS, HTTP, HTTPS, SMTP and LDAP from web, middleware and database subnets. Allow only named destinations.
  2. Use Protective DNS. Route server DNS through a resolver that can sinkhole OAST zones and log every lookup.
  3. Turn off database network functions you do not use, such as SQL Server's xp_dirtree or Oracle's UTL_HTTP, as Monograph 17 advises.
  4. Allowlist your own testers. Register approved scanner IPs and private callback domains before each test, so real alerts stand out.
  5. Block only with evidence. Keep standalone Interactsh servers on watchlists. Reserve blocks for hosts that also run C2 or show hostile callbacks. See our note on false positives.

Interactsh compared with other OAST tools

ToolWho makes itEndpoints (Monograph 17)Confirmed or suspectedWhere it runs
InteractshProjectDiscovery, open source1,245638Mostly Amazon and DigitalOcean, US and Germany
Burp CollaboratorPortSwigger, part of Burp Suite516109Mostly Amazon and DigitalOcean, US and Germany
DNSLogDNS-focused callback service338Mostly Tencent and Alibaba, China and Hong Kong

Frequently asked questions about Interactsh

What is Interactsh?

Interactsh is an open-source out-of-band application security testing (OAST) tool from ProjectDiscovery. It generates unique domains and records DNS, HTTP, SMTP, LDAP and FTP interactions with them, so testers can confirm blind vulnerabilities.

Is Interactsh malware?

No. Interactsh is a legitimate, MIT-licensed security testing tool. Attackers also use it to confirm exploits, so callbacks from your own servers to an Interactsh domain deserve investigation.

Is Interactsh legal?

Yes. Running an Interactsh server is legal. Planting its payloads in systems you are not authorized to test is not.

What is an OAST domain?

It is a domain controlled by an out-of-band testing server. Each probe gets a unique subdomain, so a lookup or request to that subdomain shows which target reacted.

Why is my server contacting an Interactsh domain?

Usually because someone sent it a probe containing that domain, and the server processed it. Find the inbound request that came just before the callback, check whether it was an approved test, and patch the flaw it reached.

How do I detect Interactsh traffic?

Alert on server-tier lookups of random labels under OAST domains, on outbound traffic from subnets that should not reach the internet, and on Interactsh server headers. Odin Research publishes Suricata, Sigma, YARA and KQL rules in its Defense Kit.

Interactsh vs Burp Collaborator: what is the difference?

Both catch out-of-band callbacks. Burp Collaborator is part of PortSwigger's Burp Suite. Interactsh is a free, open-source alternative from ProjectDiscovery that the project says was inspired by Collaborator and that integrates with Nuclei.

Should I block all Interactsh servers?

No. Most are run by testers. Odin Research sends standalone Interactsh servers to hunting and blocks only hosts with a second hostile signal, such as a C2 listener on the same address.

Sources

  1. ProjectDiscovery, Interactsh repository, GitHub. github.com/projectdiscovery/interactsh (checked 21 November 2026)
  2. PortSwigger, "What is OAST (Out-of-band Application Security Testing)?". portswigger.net/burp/application-security-testing/oast (checked 21 November 2026)
  3. Palo Alto Networks Unit 42, "Attackers Are Taking Advantage of the Open-Source Service Interactsh for Malicious Purposes", 14 October 2021. unit42.paloaltonetworks.com/exploits-interactsh/ (checked 21 November 2026)
  4. VulnCheck, "The Mystery OAST Host Behind a Regionally Focused Exploit Operation", 27 November 2025. www.vulncheck.com/blog/mystery-oast (checked 21 November 2026)
  5. NVD, CVE-2017-9506. nvd.nist.gov/vuln/detail/CVE-2017-9506 (checked 21 November 2026)
  6. NVD, CVE-2017-12629. nvd.nist.gov/vuln/detail/CVE-2017-12629 (checked 21 November 2026)

Related tools and terms

All Odin Research explanations · Brandefense glossary

Odin Research

Get the rules, not just the definition

Every Odin Research report ships with verified indicators and detection rules you can load today.

Open the Defense Kit