Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain
-
AccessGain code execution on one cluster node
-
DeliverySubmit crafted value via internal cluster interface
-
ExploitValue injected into monitoring config on second node
-
ExecutionOS command executes as unprivileged service account
-
ImpactLateral foothold across cluster nodes
Vulnerability Assessment
| Weakness |
CWE
OWASP Top 10:2025
A05:2025 - Injection
ASVS 5.0
V1 - Encoding and Sanitization CRLF sequence sanitization
|
| Exploitation Conditions | Requires a MULTI-NODE (clustered) Kiteworks Core deployment - single-node installs are not exposed via this path. The attacker must ALREADY possess code execution on one appliance node, since the vulnerable input is reachable only through the internal cluster interface, not from the public network. Exploitation writes an unvalidated value into the monitoring configuration of a second node; resulting command execution is limited to an unprivileged service account, so it does not directly yield root/privileged control on the target node. Affected versions are Kiteworks Core below 9.5.1; patched in 9.5.1. |
| Impact | This is a genuinely low-to-medium priority issue despite the OS-command-execution potential, because every strong exploitation signal is gated. The CVSS 3.1 vector (AV:L/AC:L/PR:L/UI:N/S:C/C:L/I:L/A:L, 6.3) explicitly requires LOCAL access and LOW privileges - and the description clarifies that 'local' here means the attacker must ALREADY have code execution on one appliance node in a multi-node cluster. There is no EPSS score, no CISA KEV listing, and no public POC identified at time of analysis, so exploitation status is: no public exploit identified at time of analysis. The scope change (S:C) is the one aggravating factor - the injected value crosses from one node into the monitoring configuration of a different node - but the resulting command execution is confined to an unprivileged service account, which is why C/I/A are all Low rather than High. No SSVC decision data was supplied. Net: this is a real defense-in-depth flaw for lateral movement within a compromised cluster, not a widespread remote-exploitation priority. It should be patched but does not warrant emergency response absent evidence of an initial-access chain. |
| Exploit ScenarioDerived | An attacker who has already compromised one Kiteworks appliance node (via a separate initial-access vulnerability) submits a crafted value through the internal cluster interface. That value is written into the monitoring configuration on a second node without sufficient neutralization of special/CRLF elements (CWE-93), causing OS commands to run on that second node as the unprivileged service account, extending the attacker's foothold laterally across the cluster. No public POC is identified at time of analysis. |
Get an e-mail when this CVE is exploited, patched or re-scored. Free account.
Alert meAnalysisAIv1 – Sep 30, 2026
OS command execution on a second appliance node is possible in multi-node (clustered) Kiteworks Core deployments running versions below 9.5.1, but only for an attacker who has already achieved code execution on another node in the same cluster. An unvalidated value submitted through an internal cluster interface is written into the monitoring configuration of a peer node, where it can be interpreted as an OS command running under an unprivileged service account. There is no public exploit identified at time of analysis and no confirmed active exploitation (CISA KEV); the CVSS 3.1 score of 6.3 (AV:L/PR:L/S:C) reflects the strong prerequisites - local-adjacent access, existing low-privilege foothold, and a multi-node topology.
RemediationAI
Vendor-released patch: Kiteworks Core 9.5.1, which remediates the missing neutralization on the internal cluster interface; upgrade all nodes in every cluster to 9.5.1 or later, since a mixed-version cluster leaves at least one peer writable by an older node. Consult GHSA-rpxx-jm93-w5j7 and CSAF VA-26-274-01 for the vendor's authoritative guidance and any release notes on cluster upgrade ordering. If an immediate upgrade is not feasible, the practical compensating controls follow directly from the exploitation prerequisites: (1) treat this as a post-compromise escalation path and prioritize detecting and containing any existing code execution on cluster nodes, since that is a hard prerequisite - trade-off is that this is detection, not prevention; (2) tightly segment and firewall the internal cluster/monitoring interface so only legitimate peer node addresses can reach it, with the trade-off that misconfigured rules can break cluster health checks and monitoring ingestion; (3) monitor and alert on unexpected changes to monitoring configuration files and unexpected child processes spawned by the monitoring service account, accepting that this detects rather than blocks exploitation; (4) run the monitoring and cluster services under the least-privileged service accounts available, which limits but does not eliminate the impact given execution is already confined to an unprivileged account. Because the CVSS vector requires AV:L and PR:L, there is no meaningful benefit to perimeter-only mitigation such as blocking public ports - the vulnerable interface is internal by design.
Affected VersionsENISA
Affected Packages
Source: OSV (Open Source Vulnerabilities)
Affects your stack
Affects 1 package in euvd
Checking this CVE against your own stacks and servers is part of the Basic plan and up.
Technical ContextAI · show details
The root cause class is CWE-93 (Improper Neutralization of CRLF Sequences), which in this context manifests as injection into a configuration file consumed by a downstream monitoring component. The vulnerable data path is the internal cluster interface used for node-to-node coordination in Kiteworks Core; a value passed across that interface is persisted into monitoring configuration on a peer node without adequate neutralization of control/line-break characters, so injected content is parsed as additional directives or commands when the monitoring service processes its configuration. Because the write occurs on a different node than the one the attacker controls, the CVSS vector carries Scope:Changed (S:C), pushing the base score above what a same-scope injection would merit. The vector AV:L/AC:L/PR:L/UI:N describes a local/adjacent attack requiring an existing low-privileged foothold and no user interaction. Execution is constrained to an unprivileged service account, so the flaw does not by itself grant root or administrative control of the second node, and the internal interface is not reachable from the public network. CPE data (cpe:2.3:a:kiteworks:core) and ENISA EUVD-2026-90279 both tie the issue to Kiteworks Core releases prior to 9.5.1. Tags classify the issue as an RCE affecting Core.
Know when Kiteworks gets its next vulnerability
- Watch products and get an email when something new is published
- Check this CVE against your own stack and servers
- Exploit and PoC references, threat intelligence from 22 sources
No credit card · 7-day full trial
Severity by source
AV:L/AC:L/PR:L/UI:N/S:C/C:L/I:L/A:L
AV:L and PR:L because it needs prior code exec on a node reaching an internal cluster interface; S:C for cross-node impact; C/I/A Low as execution is confined to an unprivileged service account.
Primary rating from Vendor (cisa-cg).
CVSS VectorVendor: cisa-cg
Lifecycle Timeline
1DescriptionCVE.org
In multi-node deployments, an attacker who had already obtained code execution on one appliance node could submit a value through an internal cluster interface that was written into monitoring configuration on another node without sufficient validation, potentially allowing OS commands to be executed there. Execution was limited to an unprivileged service account on that node.
Threat Intelligence Overview
How this decision was reached
The engine reads each signal below and applies a fixed list of rules from top to bottom. The first rule that matches sets the recommended action – like a statement, every line is shown.
What every action level means
SSVC AssessmentENISA
Priority Score
Attack Technique
Remote Code Execution represents the critical moment when an attacker successfully runs arbitrary code on a target system without physical access.
Unauthenticated attackers who know a target's email address can take over Kiteworks Core accounts that use locally store
Improper certificate-assignment controls in Kiteworks Email Protection Gateway allow a remote, unauthenticated attacker
Stored cross-site scripting in Kiteworks Core (all versions prior to 9.5.0) enables unauthenticated attackers to persist
Authentication bypass in Kiteworks Email Protection Gateway (versions before 9.5.0) lets unauthenticated remote attacker
Server-side request forgery in Kiteworks Email Protection Gateway (versions before 9.5.0) lets a remote, unauthenticated
Server-side request forgery in Kiteworks Email Protection Gateway before 9.5.0 allows a remote, unauthenticated sender t
Unauthenticated server-side request forgery in Kiteworks Email Protection Gateway before version 9.5.0 lets a remote att
Server-side request forgery in Kiteworks Email Protection Gateway before 9.5.0 allows a remote, unauthenticated attacker
Server-side request forgery in Kiteworks Email Protection Gateway (all versions before 9.5.0) lets a remote, unauthentic
Command injection in Kiteworks Core's internal cluster-management function allows an attacker who already has code execu
Sandbox escape in Kiteworks Core's document-conversion feature (CVE-2026-102125, CVSS 8.8) allows code already executing
Authenticated privilege escalation in Kiteworks PDN (the Accellion-line private data network) before version 9.1.0 lets
EUVD-2026-90279
NVD
CVE.org
EPSS
Vulnrichment
GitHub
Exploit-DB