N-able N-central Initial Exploitation (Able2Shell)
High severity TLP:CLEAR Part 1 of 2 Research by Rapid7 Published 18 Nov 2026
Hypothesis
An adversary is exploiting unauthenticated XXE, OGNL injection, or path traversal vulnerabilities in N-able N-central to gain initial access and execute code on RMM management servers.
Brief
The Risk to RMM Infrastructure
N-able recently disclosed several critical vulnerabilities in N-central, collectively known as Able2Shell. These flaws allow an unauthenticated adversary to perform XML External Entity (XXE) attacks, OGNL injection, and path traversal. Because Remote Monitoring and Management (RMM) servers hold high-level access to downstream client environments, they are a priority target for initial access and lateral movement. Rapid7 details these risks in their article, Critical vulnerabilities in N-able N-central prompt hosted shutdown.
How the Hunt Flows
The hunt begins by scoping the environment for vulnerable assets. The first query searches the hb_vulnerability_finding surface for hosts flagged with the specific CVEs associated with this disclosure. This step limits the search space to systems that have not yet applied the required hotfixes.
Once the vulnerable hosts are identified, the hunt pivots to hb_http_activity. The analyst looks for rare HTTP request paths targeting N-central directories such as /central/, /mms3/, or /repository/. By baselining the frequency of these paths across the estate, the hunt isolates one-off or unusual requests that indicate scanning or exploit attempts rather than legitimate management traffic.
If the adversary successfully executes code, the activity appears in hb_process_activity. The hunt monitors for shells or interpreters—such as bash, cmd.exe, or powershell.exe—spawning from the N-central service context or within its installation paths. This provides a high-fidelity signal of remote code execution (RCE).
Finally, an agent triages the findings. It correlates the presence of vulnerabilities with the identified web request patterns and suspicious process spawns. This multi-stage approach ensures that the analyst focuses on hosts with a clear path from exposure to execution.
Blind Spots and Limitations
This hunt relies on existing vulnerability data and cleartext request patterns. If the vulnerability scanner is not current, the scoping phase may miss recently exposed or unpatched hosts. Additionally, without TLS inspection, payloads for XXE or OGNL injection remain hidden within encrypted traffic. The hunt mitigates this by focusing on the process-level behavior that follows a successful exploit.
In this series
Steps
-
Query, scoping
Vulnerable N-central instances
Identify hosts that the vulnerability scanner has already flagged with N-central CVEs.
sql, reads hb_vulnerability_findingSELECT device_uid, resource_uid, cve_uid, severity, status, titleFROM hb_vulnerability_findingWHERE instr(',' || '{{cve_ids}}' || ',', ',' || cve_uid || ',') > 0What a hit looks like. A list of unique IDs for potentially vulnerable N-central servers. Silence indicates either a patched estate or that the scanner has not yet detected the vulnerabilities.
-
Query, baseline
Rare request paths on RMM servers
Identify rare HTTP request paths that may indicate scanners or exploit attempts targeting N-central management endpoints.
sql, reads hb_http_activitySELECT url_path, COUNT(DISTINCT device_hostname) AS host_count, MIN(time) AS first_seen, MAX(time) AS last_seenFROM hb_http_activityWHERE ('{{scope_hosts}}' = '' OR instr(',' || '{{scope_hosts}}' || ',', ',' || device_hostname || ',') > 0) AND (url_path LIKE '%/central/%' OR url_path LIKE '%/mms3/%' OR url_path LIKE '%/repository/%') AND time >= datetime('now', '-{{lookback_days}} days')GROUP BY url_pathHAVING host_count <= 5ORDER BY host_count ASCWhat a hit looks like. Unusual or one-off request paths targeting N-central specific directories. Common legitimate management traffic should be filtered by the host count threshold.
-
Query, detection candidate
Shell activity from N-central context
Identify shells or interpreters running from the N-central service directories, which is a high-fidelity indicator of RCE.
sql, reads hb_process_activitySELECT device_hostname, process_name, process_cmd_line, user_name, parent_process_name, timeFROM hb_process_activityWHERE ('{{scope_hosts}}' = '' OR instr(',' || '{{scope_hosts}}' || ',', ',' || device_hostname || ',') > 0) AND (LOWER(process_path) LIKE '%/nable/%' OR LOWER(process_path) LIKE '%\nable\%') AND (LOWER(process_name) IN ('sh','bash','zsh','dash','ash','cmd.exe','powershell.exe','python','perl','php','curl','wget')) AND time >= datetime('now', '-{{lookback_days}} days')What a hit looks like. Any shell or script interpreter spawning within the N-central installation or user home path. Silence strongly suggests no successful execution has occurred on the surveyed hosts during the lookback period.
-
Agent triage
Triage N-central findings
Analyze the combined evidence of exposure, scanning, and process behavior to identify compromised hosts.
-
Decision
Route on evidence
Direct the analyst to remediation if exploitation is confirmed or likely.
-
Analyst task
Remediation and forensic review
Ensure vulnerable hosts are updated and compromised hosts are handled by IR.
-
Analyst task
Close out hunt
Record hunt findings and visibility gaps.
Coverage
2 of 5 stages covered
| Stage and ATT&CK technique | Covered | How, or why not |
|---|---|---|
| N-central Exposure and Vulnerability Scanning T1190 |
Yes | vulnerable-instances-scope, rare-web-paths-prevalence |
| Unauthenticated RCE via Web Vulnerabilities T1190 T1203 |
Yes | post-exploitation-shell-activity |
| Local Privilege Escalation to Root T1068 T1548 |
Out of scope | Belongs to another part of the 'Critical vulnerabilities in N-able N-central prompt hosted shutdown' series. |
| Unauthorized Account Creation for Persistence T1136.001 |
Out of scope | Belongs to another part of the 'Critical vulnerabilities in N-able N-central prompt hosted shutdown' series. |
| Credential and Session Disclosure T1555 |
Out of scope | Belongs to another part of the 'Critical vulnerabilities in N-able N-central prompt hosted shutdown' series. |
Blind spots
- Needs hb_http_activity with decrypted payload content. Without TLS inspection, the hunt relies on request paths and post-exploitation process behavior, which may miss subtle exploit attempts that do not result in immediate shell spawns. It would answer: Are exploitation payloads like XXE or OGNL hidden within encrypted HTTPS traffic?
- Needs Recent hb_vulnerability_finding collection. If the last scan was before the disclosure or before the hotfix was released, the scoping step will provide an inaccurate list of vulnerable assets. It would answer: Is the vulnerability scanner's view of N-central current?
Parameters and data
Parameters
| Parameter | Type | Default | What it is |
|---|---|---|---|
cve_ids | list[string] | CVE-2026-125878, CVE-2026-125879, CVE-2026-125880, CVE-2026-125881, CVE-2026-125882, CVE-2026-125883, CVE-2026-91000 | CVE IDs associated with the N-central Able2Shell disclosure. |
lookback_days | number | 14 | Days of history to examine. |
scope_hosts | list[host] | none | Hostnames to focus on; if empty, later queries run across the whole estate but filter on service paths. |
Data sources
| Source | Category | Telemetry |
|---|---|---|
| Endpoint telemetry (hb_ surfaces) | endpoint | endpoint |
| Web server / proxy logs | siem | network |
Source
An open hunt.md file. It runs anywhere that reads the format.
---
analysis: This is a hunt because it correlates external vulnerability scanning results
with internal behavioral baselining of web traffic and process execution. A simple
rule on shells spawning from a service directory might be too noisy without the
context of an unauthenticated exploit attempt seen on the same host.
blind_spots:
- id: tls-inspection-gap
question: Are exploitation payloads like XXE or OGNL hidden within encrypted HTTPS
traffic?
requires: hb_http_activity with decrypted payload content
risk: Without TLS inspection, the hunt relies on request paths and post-exploitation
process behavior, which may miss subtle exploit attempts that do not result in
immediate shell spawns.
stage: unauthenticated-rce-exploitation
- id: scanner-latency
question: Is the vulnerability scanner's view of N-central current?
requires: Recent hb_vulnerability_finding collection
risk: If the last scan was before the disclosure or before the hotfix was released,
the scoping step will provide an inaccurate list of vulnerable assets.
stage: n-central-exposure-and-scanning
coverage:
- stage: n-central-exposure-and-scanning
status: covered
steps:
- vulnerable-instances-scope
- rare-web-paths-prevalence
- stage: unauthenticated-rce-exploitation
status: covered
steps:
- post-exploitation-shell-activity
- reason: Belongs to another part of the 'Critical vulnerabilities in N-able N-central
prompt hosted shutdown' series.
stage: privilege-escalation-to-root
status: out_of_scope
- reason: Belongs to another part of the 'Critical vulnerabilities in N-able N-central
prompt hosted shutdown' series.
stage: unauthorized-account-creation
status: out_of_scope
- reason: Belongs to another part of the 'Critical vulnerabilities in N-able N-central
prompt hosted shutdown' series.
stage: credential-and-session-disclosure
status: out_of_scope
guardrails:
claims: no_unsupported
evidence: citation_required
missing_data: not_benign
telemetry: untrusted
hunt:
applicability: campaign-specific
handoff: promote-to-detection
justification: N-central RMM servers are highly privileged infrastructure that can
be used to pivot into thousands of managed endpoints; protecting them from unauthenticated
RCE is a critical priority.
methodology: model-assisted
trigger: intel-report
hypothesis: An adversary is exploiting unauthenticated XXE, OGNL injection, or path
traversal vulnerabilities in N-able N-central to gain initial access and execute
code on RMM management servers.
labels:
- hunt
- attack.t1190
- attack.t1203
- credential access
- execution
- initial access
- persistence
- privilege escalation
name: N-able N-central Initial Exploitation (Able2Shell)
parameters:
cve_ids:
default:
- CVE-2026-125878
- CVE-2026-125879
- CVE-2026-125880
- CVE-2026-125881
- CVE-2026-125882
- CVE-2026-125883
- CVE-2026-91000
description: CVE IDs associated with the N-central Able2Shell disclosure.
from:
kind: article
observed: '2026-11-10'
ref: Rapid7 Able2Shell
type: list[string]
lookback_days:
default: '14'
description: Days of history to examine.
type: number
scope_hosts:
default: []
description: Hostnames to focus on; if empty, later queries run across the whole
estate but filter on service paths.
type: list[host]
provenance:
authors:
- name: Huntbase hunt generation
org: huntbase.io
generated:
by: huntbase-hunt-generation
from: https://www.rapid7.com/blog/post/etr-critical-vulnerabilities-in-n-able-n-central-prompt-hosted-shutdown
gates:
- dry-run
- lint
model: hb_google/gemini-3-flash-preview
rationale: The hunt starts with known vulnerability findings for N-central; if scanner
data is missing, the analyst should manually populate scope_hosts with the hostnames
of known N-central management servers.
references:
- name: "Rapid7 \u2014 Critical vulnerabilities in N-able N-central prompt hosted\
\ shutdown"
url: https://www.rapid7.com/blog/post/etr-critical-vulnerabilities-in-n-able-n-central-prompt-hosted-shutdown
related:
- hunt: n-central-privilege-escalation-syslog-ng
reason: The privilege escalation to root via syslog-ng is a distinct post-exploitation
step requiring different Linux audit telemetry.
relation: out-of-scope-alternative
scenario:
stages:
- name: N-central Exposure and Vulnerability Scanning
observables:
- N-able N-central instances before version 2026.4.1 (build 2026.4.1.26)
- N-central Appliance 2026.4.0.31
- HTTP scanning activity targeting N-central management interfaces
slug: n-central-exposure-and-scanning
tactic: initial-access
techniques:
- T1190
- name: Unauthenticated RCE via Web Vulnerabilities
observables:
- XML external entity (XXE) payloads in HTTP requests
- OGNL injection strings in web parameters
- Path traversal sequences (../) in file-repository upload paths
- Administrator viewing malicious MMS3 patch metadata
slug: unauthenticated-rce-exploitation
tactic: execution
techniques:
- T1190
- T1203
- name: Local Privilege Escalation to Root
observables:
- Modification of nable-writable syslog-ng configuration files
- Exploitation of insecure IPC channels
- nable service account spawning shell with root privileges
slug: privilege-escalation-to-root
tactic: privilege-escalation
techniques:
- T1068
- T1548
- name: Unauthorized Account Creation for Persistence
observables:
- New local user accounts created on the N-central appliance
- Unauthorized administrative account creation
slug: unauthorized-account-creation
tactic: persistence
techniques:
- T1136.001
- name: Credential and Session Disclosure
observables:
- Access to ProductAdmin credentials
- Exploitation of session-handling weaknesses to disclose credentials
slug: credential-and-session-disclosure
tactic: credential-access
techniques:
- T1555
summary: The Able2Shell disclosure identifies 33 vulnerabilities in the N-able N-central
RMM platform, enabling unauthenticated remote code execution (RCE) via XXE, OGNL
injection, and path traversal. Once initial access is gained, attackers can escalate
privileges to root through insecure IPC or syslog-ng configurations and maintain
persistence via unauthorized account creation, putting downstream managed environments
at risk.
series:
index: 1
slug: critical-vulnerabilities-in-n-able-n-central-prompt-hosted-shutdown
title: Critical vulnerabilities in N-able N-central prompt hosted shutdown
total: 2
severity: high
targets:
analyst:
name: Tier-2 analyst
role: analyst
endpoint:
category: endpoint
name: Endpoint telemetry (hb_ surfaces)
telemetry:
- endpoint
hunter:
agent: true
name: Hunt agent
web:
category: siem
name: Web server / proxy logs
telemetry:
- network
tlp: clear
type: investigation
---
# N-able N-central Initial Exploitation (Able2Shell)
This hunt targets the initial exploitation phase of the N-able N-central 'Able2Shell' disclosure. It focuses on identifying exposed instances that have not yet applied Hotfix 2 and monitoring for subsequent unauthenticated remote code execution. The hunt first scopes the environment for vulnerable systems using existing vulnerability findings. It then identifies rare web request patterns on those systems that might indicate scanning or exploit attempts. Finally, it looks for suspicious process behavior, such as shells or file-transfer tools spawning from the N-central service context, which corroborates successful code execution. An agent weighs these signals to identify compromised RMM infrastructure.
## vulnerable-instances-scope
<!-- Vulnerable N-central instances -->
Identify hosts that the vulnerability scanner has already flagged with N-central CVEs.
```sqlite target=endpoint role=scoping params=(cve_ids=cve_ids)
~~~yaml
expected: A list of unique IDs for potentially vulnerable N-central servers. Silence
indicates either a patched estate or that the scanner has not yet detected the vulnerabilities.
reads:
- device_uid
- resource_uid
- cve_uid
- severity
- status
- title
silence: not_evidence_of_absence
source: hb_vulnerability_finding
verified: dry-run
verified_at: '2026-11-18'
~~~
SELECT device_uid, resource_uid, cve_uid, severity, status, title FROM hb_vulnerability_finding WHERE instr(',' || '{{cve_ids}}' || ',', ',' || cve_uid || ',') > 0
```
## rare-web-paths-prevalence
<!-- Rare request paths on RMM servers -->
Identify rare HTTP request paths that may indicate scanners or exploit attempts targeting N-central management endpoints.
```sqlite target=web role=baseline params=(lookback_days=lookback_days, scope_hosts=scope_hosts)
~~~yaml
baseline:
compare: first_seen
window: '{{lookback_days}}d'
expected: Unusual or one-off request paths targeting N-central specific directories.
Common legitimate management traffic should be filtered by the host count threshold.
prevalence:
by: device_hostname
key:
- url_path
rare_below: 5
reads:
- url_path
- device_hostname
- time
silence: not_evidence_of_absence
source: hb_http_activity
verified: dry-run
verified_at: '2026-11-18'
~~~
SELECT url_path, COUNT(DISTINCT device_hostname) AS host_count, MIN(time) AS first_seen, MAX(time) AS last_seen FROM hb_http_activity WHERE ('{{scope_hosts}}' = '' OR instr(',' || '{{scope_hosts}}' || ',', ',' || device_hostname || ',') > 0) AND (url_path LIKE '%/central/%' OR url_path LIKE '%/mms3/%' OR url_path LIKE '%/repository/%') AND time >= datetime('now', '-{{lookback_days}} days') GROUP BY url_path HAVING host_count <= 5 ORDER BY host_count ASC
```
## post-exploitation-shell-activity
<!-- Shell activity from N-central context -->
Identify shells or interpreters running from the N-central service directories, which is a high-fidelity indicator of RCE.
```sqlite target=endpoint role=detection-candidate params=(lookback_days=lookback_days, scope_hosts=scope_hosts)
~~~yaml
expected: Any shell or script interpreter spawning within the N-central installation
or user home path. Silence strongly suggests no successful execution has occurred
on the surveyed hosts during the lookback period.
reads:
- device_hostname
- process_name
- process_cmd_line
- user_name
- parent_process_name
- time
- process_path
silence: evidence_of_absence
source: hb_process_activity
verified: dry-run
verified_at: '2026-11-18'
~~~
SELECT device_hostname, process_name, process_cmd_line, user_name, parent_process_name, time FROM hb_process_activity WHERE ('{{scope_hosts}}' = '' OR instr(',' || '{{scope_hosts}}' || ',', ',' || device_hostname || ',') > 0) AND (LOWER(process_path) LIKE '%/nable/%' OR LOWER(process_path) LIKE '%\nable\%') AND (LOWER(process_name) IN ('sh','bash','zsh','dash','ash','cmd.exe','powershell.exe','python','perl','php','curl','wget')) AND time >= datetime('now', '-{{lookback_days}} days')
```
## exploitation-triage
<!-- Triage N-central findings -->
```agent target=hunter
cite: required
context:
- vulnerable-instances-scope
- rare-web-paths-prevalence
- post-exploitation-shell-activity
max_iterations: 4
objective: Determine if any N-central server shows signs of successful exploitation.
Correlate vulnerable status from the scoping query with rare web paths and high-fidelity
shell activity from the behaviour query. Prioritize hosts where shell processes
spawn from the nable service context.
success_criteria: A summary citing specific hosts, their vulnerability status, and
any suspicious process command lines discovered.
tools:
- endpoint
- web
```
## exploitation-route
<!-- Route on evidence -->
if~: "the triage verdict is suspicious or malicious for at least one host" (confidence: high, judge=hunter)
then: → remediation-review
indeterminate: → remediation-review
unavailable: → remediation-review (blind_spot: tls-inspection-gap)
else: → close-out
## remediation-review
<!-- Remediation and forensic review -->
```manual target=analyst
1. Review all hosts identified as vulnerable and verify application of N-central 2026.4 Hotfix 2. 2. For any host where 'post-exploitation-shell-activity' found results, initiate incident response procedures to investigate for lateral movement and credential theft. 3. Confirm that Agents and Probes have been scheduled for upgrade.
```
→ close-out
## close-out
<!-- Close out hunt -->
```manual target=analyst
Document the number of vulnerable versus patched N-central instances. Record any identified exploitation attempts and summarize the remediation status of the environment.
```
→ end
Run this hunt against your own data.
In Huntbase, Scout runs every step across the tools you already connect, checks back through your history and returns a verdict with the evidence. The hunt.md file is open, so you can also take it anywhere that reads the format.
Drafted by machine (huntbase-hunt-generation, hb_google/gemini-3-flash-preview), checked by dry-run, lint, then reviewed by a person.