N-able N-central: Able2Shell zero-days exploited even after HF2
Over 30 public exploits against N-central, an emergency hotfix and new chains that hit patched servers too: technical details, IoCs and what to do now.

Between 8 and 9 November 2026, researcher Sean Bales published 33 exploits for N-able N-central on GitHub without notifying the vendor first. N-central is the remote monitoring and management (RMM) platform used by MSPs and IT teams. N-able shut down its cloud environments in the Americas, Europe and Asia-Pacific, told on-premises customers to take their servers offline and released a hotfix on 10 November. On 12 November two new exploits appeared that also work against patched servers, and one is already being exploited.
The risk goes beyond a single machine: whoever controls an RMM server can run commands on every managed device, including those in customer networks.
Able2Shell: the four attack chains
The package, called Able2Shell, was tested on N-central Appliance 2026.4.0.31, a pre-release build; according to the researcher, almost all the flaws also affect earlier versions. It includes authentication bypasses, credential disclosure, arbitrary file writes, directory traversal, XSS, XXE and remote code execution. The exploits combine into four chains:
- Chain 1, unauthenticated: an XXE in the service import function lets the attacker steal a trusted local session, reset an account password and run commands as the
nableservice user. The same chain creates a Default Administrator account with attacker-chosen credentials. - Chains 2 and 3: a predictable appliance passcode allows an anonymous attacker to inject malicious patch metadata. When an administrator opens the patch details, a stored XSS runs commands as
nable, or hijacks the session to submit an audit export configuration that runs commands as root. - Chain 4: a directory traversal in the software repository writes an executable page into the web folder; a passwordless installer that reads a configuration from a folder writable by
nablethen leads to root.
Hotfix 2026.4 HF2 and the seven CVEs
N-central 2026.4 Hotfix 2 (build 2026.4.1.26) fixes seven CVEs and, according to the release notes quoted by Infosecurity Magazine, temporarily disables the built-in remote control. INCIBE-CERT rates 2 of them critical and 5 high. CVSS scores reported by SecurityWeek:
- CVE-2026-125883, multiple XXE vulnerabilities (9.3);
- CVE-2026-125881, SSRF in schema validation during service import (9.3);
- CVE-2026-125878 (8.7): OGNL injection according to CSO Online and Infosecurity Magazine, XSS according to SecurityWeek;
- CVE-2026-125879, privilege escalation to root via a syslog-ng configuration writable by
nable(8.6); - CVE-2026-125882, path traversal in repository file upload leading to code execution (8.6), reported by Nate Fair of Sprocket Security;
- CVE-2026-91000, unauthenticated IPC channels leading to local privilege escalation (8.4);
- CVE-2026-125880, stored XSS in MMS3 patch metadata (8.3).
According to N-able, all flaws exploitable without authentication are fixed; the rest require an N-central account, often with elevated privileges, and will be addressed in later hotfixes. The company has not explained how the seven CVEs map to the 33 exploits, which remain public. The cloud environments are back online; after the upgrade, Agents and Probes must be updated too, both in the cloud and on-premises.
4EverAble2Shell: exploited on HF2 servers
On 12 November Bales released two exploits tested precisely on 2026.4.1.26 HF2. 4EverAble2Shell gives unauthenticated root access, but only with the Legacy API enabled: it abuses an Apache Axis flaw (CVE-2026-90436) to run commands as nable, then, according to Previdian, uses the privileged setLDFileOwner helper and a symbolic link to make a root-owned script writable and launches it through serverControl. Agent2Shell targets the agent and, according to the researcher, grants SYSTEM privileges on managed Windows devices.
Exploitation began within hours. Previdian and Huntback saw the same address probing their honeypots around 08:30 UTC and attempting the full chain from 09:16 UTC. The payload downloaded a script from 157.20.241[.]16 and installed the XMRig miner, capped at about half of the CPU cores. Using the mining pool statistics, at 12:50 UTC Huntback counted 21 compromised N-central servers, 19 of them on HF2, in the United States, Canada, South Africa, the United Kingdom, Germany, France and Sweden. N-able disabled the Intune integration as a precaution, restricted some features in the cloud environments and asks customers to make sure the Legacy API is off, as it is by default.
MITRE ATT&CK techniques: T1190 (exploit public-facing application), T1068 (privilege escalation), T1136 (create account), T1496 (resource hijacking) and, for MSP customers, T1199 (trusted relationship).
Impact for European companies
Belgium’s CCB and INCIBE-CERT published advisories, and compromised servers include some in Germany, France and Sweden. Censys counted 1,728 N-central instances exposed to the internet. MSPs fall under the NIS2 directive (ICT service management): the compromise of an RMM server can become a significant incident to be notified, for the MSP and for its customers in covered sectors, healthcare included.
What to do now
- Upgrade on-premises servers to 2026.4 HF2, then Agents and Probes; install the next hotfixes as soon as they ship, because HF2 does not stop 4EverAble2Shell.
- Check that the Legacy API is disabled and restrict access to the N-central console to authorised addresses.
- Search logs for requests to
/dms/services/ServerUI, especially withjline.internal.TerminalLineSettings; look forxvd.servicesystemd services, files in/tmp/.xvdor/var/tmp/.xvd, mining processes and connections to 157.20.241[.]16. If you find them, treat the server as fully compromised at root level. - Huntress advises reviewing authentication activity since 8 November, looking for administrator accounts, scripts and remote sessions you did not create, and assessing every managed customer network, not just the server. The version number may not change after the fix: verify it using N-able’s guidance.
- ThreatDefence suggests preserving logs and snapshots before making changes and, once the server is restored, rotating credentials for administrators and technicians, asset discovery databases, agent registration tokens and agent keys.
How ARXDOME helps
ARX Expose shows which of your company’s services are visible from the internet, including management consoles. ARX Sentry collects logs and events from monitored servers, useful for hunting anomalous requests and processes like those described, and ARX Trace lets analysts reconstruct what happened across multiple machines.
Last updated:


