EUVD-2026-85168
GHSA-q25p-jh2x-hv2j
Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain
-
AccessHost malicious web page
-
DeliveryLure victim running CGServiSign to visit
-
ExploitBrowser reaches local service interface
-
ExecutionInject OS commands via interface
-
ImpactExecute arbitrary commands on victim host
Vulnerability Assessment
| Weakness |
CWE
OWASP Top 10:2025
A05:2025 - Injection
ASVS 5.0
V1 - Encoding and Sanitization OS command sanitization, avoid shell exec
MITRE ATT&CK
2 techniques across 1 tactic: Execution The technique list with ATT&CK links and the kill-chain view is part of the Basic plan and up. |
| Exploitation Conditions | Exploitation requires that the victim has Changing CGServiSign installed and its local service interface running/listening on their machine, and that the victim actively visits an attacker-controlled web page (UI:A/UI:R) while that service is reachable from the browser. No authentication is needed (PR:N) and attack complexity is low (AC:L), but the dependence on user interaction plus the requirement that this specific Taiwanese e-signature client be installed and active substantially narrows the realistic target population beyond the raw 8.6 score. |
| Impact | This is a genuine high-priority issue where the vendor CVSS 4.0 score (8.6) and the technical facts align. The vector CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:A/VC:H/VI:H/VA:H describes the classic 'localhost helper service' attack: CGServiSign runs a local service interface (bound to the victim's machine) that a malicious web page can reach from the browser, and because PR:N there is no authentication barrier, while full VC:H/VI:H/VA:H reflects arbitrary OS command execution on the victim's computer. The single meaningful limiter is UI:A/UI:R - the victim must actively browse to attacker-controlled content while CGServiSign is running. No EPSS score, no CISA KEV listing, and no public exploit code were supplied, so there is no public exploit identified at time of analysis and no evidence of active exploitation; the risk rating rests on the impact and low attack complexity, not on observed weaponization. The main uncertainty is deployment footprint: CGServiSign is a Taiwan-specific digital-signature client (TWCERT-reported), so real-world exposure is concentrated rather than widespread. No conflicting signals between the vector and the description were found. |
| Exploit ScenarioDerived | An attacker registers a malicious web page (or injects script into a site) that issues requests to the CGServiSign local service interface listening on the victim's machine. When a victim who has CGServiSign installed and running visits the page, the crafted request injects OS commands that execute in the victim's local context, giving the attacker code execution on that endpoint. No authentication is required and no public PoC was provided in the source data. |
Get an e-mail when this CVE is exploited, patched or re-scored. Free account.
Alert meAnalysisAIv1 – Sep 23, 2026
Unauthenticated OS command injection in Changing CGServiSign enables a remote attacker to run arbitrary operating-system commands on a victim's machine when that victim is lured into visiting a malicious web page while CGServiSign's local service interface is running. The attacker needs no credentials and no access to the host (CVSS 4.0 8.6, PR:N, AC:L), but exploitation depends on user interaction (UI:A) plus the presence of this specific Taiwanese e-signature client with its local listener active, which meaningfully narrows the realistic target set. No public exploit code and no confirmed active exploitation were identified at time of analysis, and because the CPE is a wildcard over all versions, every deployment must be considered affected until a fixed build is published.
RemediationAI
No vendor-released patch was identified at time of analysis, so remediation must start with compensating controls while tracking the TWCERT advisories (https://www.twcert.org.tw/tw/cp-132-11213-28a81-1.html and https://www.twcert.org.tw/en/cp-139-11214-89281-2.html) for a fixed build; the wildcard CPE does not name a corrected version, so do not assume any upgrade resolves the issue until the vendor confirms it. The most effective immediate mitigation is to stop or disable the CGServiSign local service/agent when signing workflows are not in use, which removes the attack surface entirely but breaks on-demand signing and smart-card integration until it is restarted - a trade-off that is acceptable for endpoints that rarely sign. Where the service must stay available, restrict who can reach the loopback listener and prevent browser-originated cross-origin requests to it (for example by blocking the service's local port through host firewall rules scoped to the browser binaries, or by using a browser policy/extension that denies requests to localhost endpoints), accepting that some legitimate web-to-client signing flows may break and require reconfiguration. On managed fleets, treat the client as a high-risk local agent: remove it from users who do not need it, keep browsers patched and enforce strict egress filtering so drive-by pages cannot be reached, and monitor host telemetry for unexpected child processes spawned by the CGServiSign service (cmd.exe, powershell.exe, or shell utilities) as a detection for exploitation attempts. Do not rely on network-perimeter controls alone, since exploitation originates from the victim's browser and terminates on the victim's own machine.
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 is CWE-78 OS Command Injection: the CGServiSign local service exposes an interface (typically an HTTP/loopback listener used for browser-client integration in e-signature and smart-card workflows) that constructs shell commands from input without adequate neutralization, so an attacker who can reach that interface can append or inject commands that execute with the privileges of the service. Because the listener binds locally and is reachable from a browser running on the same host, a drive-by web page can trigger the vulnerable endpoint cross-origin, turning a client-side web visit into local code execution - a pattern common to local-agent helpers in signing, banking, and HSM/PKI middleware. The supplied CVSS 4.0 vector (AV:N/AC:L/AT:N/PR:N/UI:A/VC:H/VI:H/VA:H) confirms remote reachability, low complexity, no authentication, and active user participation, with all three confidentiality, integrity, and availability impacts rated High. The follow-on scope metrics are all Negated (SC:N/SI:N/SA:N), so the assessed impact is contained to the victim's own machine rather than pivoting to other systems, which aligns with our independent CVSS 3.1 assessment (AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H). The CPE cpe:2.3:a:changing:cgservisign:*:*:*:*:*:*:*:* carries no version bounds, so no version is exonerated by the available data.
Know when Cgservisign 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
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:A/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N
Delivered over the network via a web page (AV:N) with no auth (PR:N) but requiring the victim to visit it (UI:R); commands run locally on the same host (S:U) with full C/I/A impact.
Primary rating from Vendor (twcert).
CVSS VectorVendor: twcert
DescriptionCVE.org
CGServiSign developed by Changing has a OS Command Injection vulnerability. Unauthenticated remote attackers can induce victims to visit a malicious web page and inject arbitrary OS commands through the local service interface, resulting in command execution on the victim's local computer.
Executive ViewAI
All versions of the Changing CGServiSign e-signature client expose a locally running service interface that an unauthenticated remote attacker can abuse to execute arbitrary operating-system commands on an end-user workstation, requiring only that the user be lured to a malicious web page while the client is active. Because the affected software base is a wildcard across all versions and no vendor fix has been released, every deployment of this client in the organization should be treated as vulnerable until a patched build is published. Leadership should care because a single successful exploitation yields full control of the user's machine, putting corporate credentials, documents, and e-signature integrity at risk and providing a foothold for lateral movement into higher-value systems.
No vendor-released patch identified at time of analysis. Interim mitigations: stop or disable the CGServiSign local service interface where the client is not essential; block inbound network access to that local listener via host firewall except from localhost; isolate affected endpoints on a restricted segment; enforce strict web browsing and phishing controls to reduce the user-interaction vector; monitor endpoint telemetry for command-shell processes spawned by CGServiSign; and plan migration to an alternative or vendor-fixed e-signature client.
Threat Intelligence Overview
NVD
CVE.org
EPSS
Vulnrichment
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
OS command injection occurs when an application passes unsanitized user input directly into system shell commands.
Exploit-DB