Fortinet FortiSandbox CVE-2026-25089 detection
| Reporter | Title | Published | Views | Family All 23 |
|---|---|---|---|---|
| The vulnerability of the graphical user interface of Fortinet’s FortiSandbox and FortiSandbox Cloud systems allows a perpetrator to execute arbitrary commands through specially crafted HTTP requests. | 17 Jun 202600:00 | – | bdu_fstec | |
| CVE-2026-25089 | 10 Jun 202605:00 | – | circl | |
| Fortinet FortiSandbox OS Command Injection Vulnerability | 16 Jul 202600:00 | – | cisa_kev | |
| Fortinet多款产品 操作系统命令注入漏洞 | 9 Jun 202600:00 | – | cnnvd | |
| Fortinet FortiSandbox OS Command Injection Vulnerability (CNVD-2026-25587) | 11 Jun 202600:00 | – | cnvd | |
| CVE-2026-25089 | 9 Jun 202614:27 | – | cve | |
| CVE-2026-25089 | 9 Jun 202614:27 | – | cvelist | |
| EUVD-2026-35443 | 9 Jun 202618:30 | – | euvd | |
| Second-Order OS Command Injection via JSON Input on start vnc feature | 9 Jun 202600:00 | – | fortinet | |
| Fortinet FortiSandbox 4.4.x < 4.4.9 / 5.0.x < 5.0.6 Multiple Vulnerabilities | 3 Jul 202600:00 | – | nessus |
10
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
import re
from typing import Optional, Tuple
from kittysploit import *
from lib.protocols.http.http_client import Http_client
class Module(Scanner, Http_client):
__info__ = {
"name": "Fortinet FortiSandbox CVE-2026-25089 detection",
"description": (
"Fingerprints FortiSandbox and flags versions affected by CVE-2026-25089 "
"(unauthenticated OS command injection in the VNC start handler): "
"5.0.0–5.0.5, 4.4.0–4.4.8, and all 4.2.x."
),
"author": ["KittySploit Team"],
"severity": "critical",
"cve": "CVE-2026-25089",
"references": [
"https://nvd.nist.gov/vuln/detail/CVE-2026-25089",
"https://www.fortiguard.com/psirt/FG-IR-26-141",
],
"modules": [
"exploits/linux/http/fortinet_fortisandbox_cve_2026_25089_rce",
],
"tags": [
"web",
"scanner",
"fortinet",
"fortisandbox",
"command-injection",
"rce",
"cve-2026-25089",
],
"agent": {
"risk": "active",
"effects": ["network_probe"],
"expected_requests": 2,
"reversible": True,
"approval_required": False,
"produces": ["tech_hints", "risk_signals", "endpoints"],
"cost": 1.0,
"noise": 0.2,
"value": 1.0,
"requires": {
"min_endpoints": 0,
"min_params": 0,
"tech_hints_any": ["fortinet", "fortisandbox"],
"tech_hints_all": [],
"specializations_any": [],
"risk_signals_any": [],
"auth_session": False,
"capabilities_any": [],
"capabilities_all": [],
"confidence_min": {},
"confidence_min_any": {},
"endpoint_pattern_any": [],
"param_any": [],
"api_surface_ready": False,
},
"chain": {
"produces_capabilities": [
{"capability": "rce", "from_detail": ""},
],
"consumes_capabilities": [],
"option_bindings": {},
"suggested_followups": [
"exploits/linux/http/fortinet_fortisandbox_cve_2026_25089_rce",
],
},
},
}
port = OptPort(443, "FortiSandbox HTTPS port", True)
ssl = OptBool(True, "Use HTTPS", True, advanced=True)
_VERSION_PATTERNS = (
re.compile(r"fortisandbox[^0-9]{0,24}(5\.\d+\.\d+)", re.I),
re.compile(r"fortisandbox[^0-9]{0,24}(4\.\d+\.\d+)", re.I),
re.compile(r'"version"\s*:\s*"(5\.\d+\.\d+)"', re.I),
re.compile(r'"version"\s*:\s*"(4\.\d+\.\d+)"', re.I),
re.compile(r"FortiSandbox\s+v?(5\.\d+\.\d+)", re.I),
re.compile(r"FortiSandbox\s+v?(4\.\d+\.\d+)", re.I),
)
_PROBE_PATHS = ("/", "/login")
@staticmethod
def _version_tuple(value: str) -> Tuple[int, ...]:
parts = []
for token in re.findall(r"\d+", value or ""):
parts.append(int(token))
while len(parts) < 3:
parts.append(0)
return tuple(parts[:3])
def _parse_version(self, text: str) -> str:
for pattern in self._VERSION_PATTERNS:
match = pattern.search(text or "")
if match:
return match.group(1).strip()
return ""
def _is_vulnerable(self, version: str) -> Optional[bool]:
if not version:
return None
major, minor, patch = self._version_tuple(version)
if major == 5 and minor == 0:
return patch <= 5
if major == 4 and minor == 4:
return patch <= 8
if major == 4 and minor == 2:
return True
return False
@staticmethod
def _looks_like_fortisandbox(body: str, headers: dict) -> bool:
text = (body or "").lower()
if "fortisandbox" in text or "forti sandbox" in text:
return True
blob = " ".join(f"{k}:{v}" for k, v in (headers or {}).items()).lower()
return "fortinet" in blob and "sandbox" in text
def run(self):
try:
detected = False
version = ""
evidence_path = ""
for path in self._PROBE_PATHS:
response = self.http_request(
method="GET",
path=path,
allow_redirects=True,
timeout=max(int(self.timeout or 10), 10),
)
if not response or response.status_code != 200:
continue
body = response.text or ""
headers = dict(response.headers)
if self._looks_like_fortisandbox(body, headers):
detected = True
evidence_path = path
parsed = self._parse_version(body)
if parsed:
version = parsed
evidence_path = evidence_path or path
if detected and version:
break
if not detected:
return False
if not version:
self.set_info(
severity="info",
cve="CVE-2026-25089",
reason="FortiSandbox detected but version could not be extracted",
path=evidence_path,
)
return True
if self._is_vulnerable(version):
self.set_info(
severity="critical",
cve="CVE-2026-25089",
reason=(
f"FortiSandbox {version} detected at {evidence_path}; "
"within CVE-2026-25089 affected range"
),
version=version,
path=evidence_path,
confidence="high",
)
return True
self.set_info(
severity="info",
reason=(
f"FortiSandbox {version} detected at {evidence_path}; "
"appears patched for CVE-2026-25089"
),
version=version,
path=evidence_path,
)
return False
except Exception as exc:
print_error(f"Scanner failed: {exc}")
return False
Data
Build on a solid foundation with Vulners data
We provide the essential building blocks for cybersecurity solutions with comprehensive, structured, and constantly updated vulnerability and exploits data
Api
Power your application with Vulners API
The Vulners REST API offers reliable, high-performance access to vulnerability intelligence, with 99.9% SLA uptime and CDN-backed data delivery for seamless global access
App
Assess and manage vulnerabilities with Vulners tools
Built on top of Vulners' database and SDK, end-user solutions give security professionals and developers lightweight and powerful tools for vulnerability remediation
10 Jul 2026 11:31Current
8.8High risk
Vulners AI Score8.8
CVSS 3.19.8
EPSS0.76112
SSVC