Information Disclosure

Improper authorization in GitLab CE/EE exposes confidential issue content to any authenticated user on public projects across three active release branches. The flaw (CWE-288) allows a low-privileged, authenticated user - including users with no project membership - to read issue content that project owners explicitly marked confidential and restricted. With patch versions released on May 13, 2026 and no public exploit or KEV listing identified at time of analysis, the immediate threat is bounded, though the confidentiality impact is rated High by CVSS given the potential for sensitive business or security-relevant issue data to be disclosed.

GitLab's Jira integration exposes out-of-scope Jira issues to authenticated GitLab users across all editions (CE and EE) from version 13.7 through the patched releases, due to the integration's project-scope filter operating only as a UI display control rather than an enforced access boundary. The Changed scope (S:C) in the CVSS vector reflects that impact crosses into Atlassian Jira - a component outside GitLab's own trust domain - allowing confidential Jira issue data to leak beyond intended project boundaries. No public exploit exists and no active exploitation has been confirmed; EPSS is negligible at 0.01% (3rd percentile), placing this firmly in low-exploitation-probability territory despite a broad version range spanning over five years of releases.

OAuth scope enforcement bypass in GitLab CE/EE allows an authenticated user holding a read_api-scoped OAuth token to perform unauthorized write operations - specifically creating issues and adding comments - in private projects. Affected are all GitLab Community and Enterprise Edition installations from version 16.0 up to the patched releases 18.9.7, 18.10.6, and 18.11.3. A publicly available proof-of-concept exists, and while EPSS probability is extremely low (0.01%, 1st percentile) with no CISA KEV listing, the integrity impact is rated High given the ability to inject content into private project issue trackers.

Cleartext transmission of sensitive information in KDDI's Android parental-control app 'あんしんフィルター for au' (Anshin Filter for au) exposes communications to interception and modification by man-in-the-middle attackers. Versions prior to 4.9_b0003 transmit data over unencrypted channels, enabling an attacker with a privileged network position to read sensitive user data or tamper with app communications. Disclosed by JPCERT/CC (JVN#24167657), no public exploit code exists and CISA has not listed this in KEV; EPSS of 0.01% (3rd percentile) and SSVC exploitation status of 'none' confirm negligible current threat activity.

Cleartext transmission in Foscam VD1 Video Doorbell (firmware before V5.3.13_1072) exposes Session Description Protocol (SDP) credentials and ICE candidates over unencrypted network channels, enabling on-path attackers to intercept media stream authentication tokens, hijack real-time video/audio feeds, and abuse Foscam's TURN relay infrastructure for unauthorized traffic routing. EPSS score of 0.02% (5th percentile) suggests low widespread exploitation likelihood, though the network-accessible attack vector (AV:N) with no authentication requirement (PR:N) and low complexity (AC:L) creates risk in residential deployment scenarios where LAN or ISP-level interception is feasible.

Brute-force lockout bypass in OPNsense prior to 26.1.7 allows unauthenticated remote attackers to indefinitely circumvent the authentication failure counter, enabling unlimited credential guessing against any network-accessible login endpoint. The flaw resides in the lockout_handler logic, which interprets attacker-controlled username strings containing the keywords 'Accepted' or 'Successful login' as success signals and resets the IP-based failure counter. A publicly available proof-of-concept exploit exists (SSVC exploitation: poc), the attack is classified as automatable with no prerequisites beyond network reach, and no active exploitation is confirmed in CISA KEV. EPSS is low at 0.03% (10th percentile), suggesting limited observed exploitation at time of analysis.

Insufficient input validation in MISP's Collections model allows authenticated low-privileged users to inject malformed UUID values into Collection records, potentially causing data integrity issues or unexpected behavior in downstream code paths that assume RFC 4122-compliant UUIDs. Affected deployments are all MISP instances prior to version 2.5.37. No public exploit code exists and CISA's SSVC framework rates exploitation as none, making this a low-urgency integrity issue rather than an active threat, though integrity of Collection relationships and UUID-dependent logic is at risk until patched.

Arbitrary file read in Grafana OSS exposes server filesystem contents to authenticated low-privilege users when the sqlExpressions feature toggle is enabled. Affected versions span the 11.6.x, 12.x, and 13.0.x release trains, with fixed security builds available across all affected branches. No public exploit code exists and CISA has not added this to the Known Exploited Vulnerabilities catalog; however, the confidentiality impact is rated High by CVSS due to the potential for unrestricted file disclosure from the Grafana server's filesystem.

Improper certificate validation in Palo Alto Networks Prisma Access Agent versions below 26.2.1 on Android and Chrome OS exposes VPN sessions to man-in-the-middle interception by adjacent-network attackers. An attacker co-located on the same network segment can present any certificate for any domain issued by a trusted Certificate Authority - bypassing domain-specific validation - to intercept VPN tunnel traffic and capture sensitive device information. No public exploit code exists and no active exploitation has been confirmed (EPSS 0.00%, not in CISA KEV); however, a vendor patch is available at version 26.2.1 and should be prioritized for mobile and Chrome OS deployments.

SQL injection in Palo Alto Networks Trust Protection Foundation exposes the platform database to arbitrary command execution by authenticated attackers on an adjacent network. Affected versions span four active release trains (24.1.x, 24.3.x, 25.1.x, 25.3.x), with successful exploitation enabling full database read/write access and privilege escalation to administrative control. No public exploit code exists and no active exploitation has been identified; EPSS sits at 0.01% and SSVC exploitation status is 'none', but the SSVC technical impact rating of 'total' underscores the severity of a successful attack.

Improper certificate validation in Palo Alto Networks Prisma SD-WAN ION allows an adjacent-network attacker positioned as a man-in-the-middle to impersonate the SD-WAN controller, achieving high confidentiality, integrity, and availability impact against the vulnerable device. Affected versions span three active release trains (6.3.x, 6.4.x, and 6.5.x), with fixed builds now available from Palo Alto Networks. No public exploit identified at time of analysis, and EPSS sits at 0.00%, but SSVC rates technical impact as total given the controller-impersonation capability.

Prisma Access Agent on Windows and macOS exposes sensitive configuration data and credentials to local low-privileged users through multiple information disclosure weaknesses. Palo Alto Networks has confirmed these vulnerabilities affect agent versions prior to 26.2.1; Linux, ChromeOS, Android, and iOS deployments are explicitly out of scope. No public exploit has been identified at time of analysis, and SSVC classifies exploitation status as none, making this a low-urgency but real credential-hygiene risk on affected desktop platforms.

Sensitive vault credential disclosure in Palo Alto Networks Trust Protection Foundation allows an authenticated, adjacent-network attacker with low privileges to read secrets from the server's underlying vault (indicated by the HashiCorp tag), then leverage those secrets to impersonate any user in the environment and arbitrarily modify platform configuration. Affected across four active release branches (24.1.x, 24.3.x, 25.1.x, 25.3.x), with fixed versions available from the vendor. No public exploit code exists and CISA has not listed this in KEV; EPSS exploitation probability is 0.01%, reflecting low current threat activity.

Session impersonation in Arqit Symmetric Key Agreement Platform (SKA-Platform) prior to version 26.03 is enabled by improper idle timeout enforcement in the embedded Keycloak authentication interface, allowing browser sessions to persist beyond their intended expiry window. An attacker with physical access to an unattended device where a tenant user has an open, authenticated session can exploit the unexpired session to impersonate that user. No public exploit code exists and CISA has not listed this in KEV; EPSS is at the 0th percentile, consistent with SSVC's 'none' exploitation status and the severe physical-access prerequisite.

Unauthenticated information disclosure in Palo Alto Networks' Chronosphere Chronocollector exposes sensitive data to any attacker with adjacent network access to the collector service endpoint. All versions from 0.0.0 up to v0.116.0 are affected, with the fix released in v0.116.0. No public exploit code has been identified at time of analysis and CISA has not added this to the KEV catalog, but the CVSS 4.0 Value Density modifier of 'Concentrated' (V:C) suggests the data at risk may include high-value material such as credentials or configuration secrets.

Authenticated file read and delete in Palo Alto Networks WildFire WF-500 and WF-500-B on-premise sandboxing appliances exposes sensitive system information and permits arbitrary file deletion by low-privileged network users. The root cause is CWE-73 (External Control of File Name or Path), where user-supplied input is insufficiently validated before being used to construct file system paths. Exploitation is constrained to appliances running in the default non-FIPS configuration mode; organizations relying solely on the WildFire Public cloud service are not impacted. No public exploit has been identified at time of analysis, consistent with EPSS at 0.05% (16th percentile) and SSVC exploitation status of 'none'.

Information disclosure in Zoom Workplace for iOS before version 7.0.0 allows an authenticated user with physical access to the device to read limited confidential data by exploiting a failed protection mechanism. The vulnerability is constrained by both physical proximity and high privilege requirements, yielding a CVSS score of 1.8 - among the lowest possible. No active exploitation has been identified, and Zoom has released a fix in version 7.0.0 per their security bulletin ZSB-26006.

go-billy versions prior to v5.9.0 and v6.0.0-alpha.1 lack proper depth and cycle detection in symlink resolution, allowing authenticated remote attackers to trigger infinite loops, uncontrolled recursion, or excessive resource consumption through crafted or malformed repository data and filesystem structures. The vulnerability stems from insufficient validation and missing safety mechanisms such as cycle detection, recursion limits, or defensive handling of unexpected states. CVSS 6.5 (availability impact) reflects the authenticated requirement (PR:L) and network attack vector, with no public exploit currently identified.

State management failure in the Linux kernel liveupdate subsystem allows a local low-privileged user to trigger a use-after-free or double-free condition by retrying a failed LIVEUPDATE_SESSION_RETRIEVE_FD ioctl, resulting in kernel crash or WARN and full availability loss. The luo_file struct's retrieve status is never recorded on failure, leaving the kernel's serialization state machine inconsistent; a retry re-enters retrieve logic against partially freed data structures such as kho folio mappings. No public exploit exists and EPSS is 0.02%, placing this firmly in low-priority territory absent active use of the liveupdate feature.

Interrupt storm vulnerability in the Linux kernel xHCI USB host controller driver allows a local user to cause a complete system availability loss by triggering a Host Controller Error (HCE) via UAS storage device hotplug events. Affected kernel versions prior to 6.6.130, 6.12.78, 6.18.19, 6.19.9, and 7.0 fail to invoke xhci_halt() when HCE is detected in xhci_irq(), leaving the interrupt uncleared and the controller running - resulting in a continuous interrupt storm. No public exploit has been identified and EPSS is 0.02% (5th percentile), consistent with the hardware-interaction-dependent trigger, but availability impact is high on affected hosts where USB ports are physically accessible.

Random system freeze vulnerability in the Linux kernel's libata-core subsystem affects any Linux host physically equipped with a Seagate ST1000DM010-2EP102 BarraCuda hard drive, where the kernel's default enablement of SATA Link Power Management (LPM) causes the drive to fail its wake sequence and hang the system irrecoverably. The ST1000DM010-2EP102 belongs to the same BarraCuda family as the ST2000DM008-2FR102, for which an LPM quirk already exists in libata - the fix extends that quirk to cover this model. No active exploitation has been identified (absent from CISA KEV), and EPSS at 0.02% reflects negligible adversarial interest, consistent with this being a hardware compatibility defect rather than a traditional security flaw.

Spurious WARN_ON assertions in the Linux kernel's nouveau/gsp ACPI probe routines trigger frequently during normal NVIDIA GPU initialization, creating an availability risk on affected systems. On kernels configured with panic_on_warn=1, each triggered WARN_ON escalates to a full kernel panic, while on default configurations it produces excessive kernel log noise that degrades observability. Affected kernel versions range from the introduction of commit 176fdcbddfd2 through stable branches fixed in 6.18.19, 6.19.9, and 7.0; no active exploitation is identified (EPSS 0.02%, not in CISA KEV).

Concurrent bitfield RMW (Read-Modify-Write) corruption in the Linux kernel MMC core subsystem allows a local authenticated user to trigger spurious WARN_ON panics or system instability on MMC-enabled devices. The root cause is that host->claimed and retune control flags shared a single bitfield word; under concurrent access from __mmc_claim_host() and mmc_mq_queue_rq(), non-atomic RMW operations allow one thread's write to silently overwrite bits modified by another, corrupting MMC host state. No public exploit has been identified at time of analysis, and EPSS sits at 0.02% (7th percentile), reflecting the narrow, race-window-dependent trigger and low attacker leverage on commodity systems.

CR8 write interception mismanagement in KVM's AMD SVM implementation crashes Windows guests on AMD hypervisors with AVIC enabled. When KVM emulates INIT→WFS sequences while AVIC is temporarily deactivated, the CR8 write intercept flag is not cleared upon AVIC reactivation, leaving it permanently enabled. In isolation this is a performance regression, but combined with the TPR synchronization flaw addressed by commit d02e48830e3f, the divergence between hardware-visible and guest-visible TPR values becomes fatal to Windows guests. No public exploit identified at time of analysis; EPSS is 0.02% (7th percentile) and no CISA KEV listing exists.

Kernel availability disruption in the Linux lan78xx USB-to-Ethernet driver allows a local user with low privileges to trigger a kernel WARN in __netif_napi_del_locked() by disconnecting an affected USB Ethernet adapter. The root cause is a redundant netif_napi_del() call in the driver's disconnect path while NAPI processing is still active, conflicting with the teardown that unregister_netdev() performs automatically. No public exploit exists and EPSS is 0.02% (4th percentile), indicating very low real-world exploitation interest; this is primarily a stability and denial-of-service concern on embedded and desktop Linux systems using the SMSC/Microchip LAN78xx adapter family.

System hang via Machine Check Error (MCE) in the Linux kernel's Intel i915 DRM driver affects ICL (Ice Lake) generation GPU hardware when VRR timing registers are written before TRANS_DDI_FUNC_CTL is enabled, violating Intel BSpec 22243. Local low-privilege users on ICL systems - particularly those with external displays connected through USB-C docks experiencing link training failure - can trigger an unrecoverable system hang. No public exploit exists and EPSS is 0.02%, consistent with the hardware-specific, condition-dependent nature of the bug; no active exploitation is confirmed.

aria2c fails to properly validate Extended Key Usage (EKU) constraints in TLS server certificates, allowing attackers who possess a compromised certificate issued for a different purpose to impersonate legitimate servers. This undermines certificate-based authentication and enables man-in-the-middle attacks against aria2c downloads over HTTPS, potentially leading to delivery of malicious files or interception of sensitive data.

Uncontrolled recursion in protobufjs versions prior to 7.5.8 and 8.2.0 allows remote attackers to exhaust the JavaScript call stack by providing crafted JSON descriptors with deeply nested namespace definitions to Root.fromJSON() or Namespace.addJSON(), causing a denial of service. The vulnerability requires only network access and no authentication, though exploitation depends on the application parsing untrusted protobuf JSON descriptors.

Ecommerce Systempay 1.0 contains a weak cryptographic implementation vulnerability that allows attackers to brute force the 16-character production secret key used for payment signature generation. Rated critical severity (CVSS 9.3), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Public exploit code available and no vendor patch available.

Heap buffer over-read in NGINX's ngx_http_charset_module allows unauthenticated remote attackers to leak sensitive memory or crash worker processes when specific configuration directives (charset, source_charset, charset_map, and proxy_pass with buffering disabled) are combined. The vulnerability requires attacker-controlled conditions that depend on factors outside the attacker's control, limiting exploitability but creating real risk for affected deployments. CVSS 4.8 reflects the conditional nature of exploitation and limited scope of impact (information disclosure or availability).

Heap-use-after-free in NGINX Plus and NGINX Open Source allows unauthenticated remote attackers to trigger memory corruption in the worker process when ssl_verify_client is set to 'on' or 'optional' and ssl_ocsp is configured with a resolver. Exploitation can cause limited information disclosure or worker process restart, with CVSS 4.8 reflecting moderate impact constrained by high attack complexity. No public exploit code or active exploitation has been identified at time of analysis.

F5 BIG-IP iControl REST API allows authenticated attackers to enumerate local user account names through undisclosed requests, leading to information disclosure of administrative user identities. The vulnerability requires valid authentication credentials and network access to the iControl REST interface, affecting systems with BIG-IP versions that have not reached End of Technical Support. CVSS 4.3 (low) reflects the requirement for prior authentication and confidentiality-only impact, though the enumeration of administrative accounts could facilitate downstream attacks.

Authenticated high-privilege attackers with Resource Administrator or Administrator roles can download sensitive files from F5 BIG-IP iControl SOAP interface due to improper path validation. The vulnerability requires valid administrative credentials and does not affect versions that have reached End of Technical Support, limiting exposure to actively maintained deployments. No public exploit code or active exploitation has been identified.

F5 BIG-IP DNS when provisioned contains an undisclosed TMOS Shell (tmsh) command vulnerability allowing highly privileged authenticated attackers to view sensitive information. The vulnerability requires high-privilege account access and local shell access (AV:L, PR:H), limiting real-world exploitation to insider threats or post-compromise scenarios where an attacker has already obtained administrative credentials on the management interface.

Denial of service in F5 BIG-IP affects the Traffic Management Microkernel (TMM) when Bidirectional Forwarding Detection (BFD) is configured with static or dynamic routing protocols. Undisclosed traffic patterns cause TMM to stop processing BFD packets, triggering unintended failover of the configured routing protocol. Remote unauthenticated attackers can trigger this condition over the network with low complexity, resulting in availability loss for BFD-dependent routing operations.

BIG-IP DNS provisioning exposes SSH passwords in cleartext within iControl REST API responses and audit logs when using the gtm_add and bigip_add commands, allowing highly privileged authenticated attackers with audit log access to retrieve sensitive credentials. The vulnerability affects all supported BIG-IP DNS versions and carries a CVSS score of 4.4 with low real-world exploitation risk due to the requirement for local access and high privilege level.

Sensitive information disclosure in F5 BIG-IP and BIG-IQ allows authenticated administrators with resource administrator role to view confidential data via undisclosed iControl REST endpoints or TMOS Shell commands. The vulnerability requires high-privilege authentication and produces no system modification or availability impact, limiting real-world risk despite network accessibility. Vendor has released patches addressing the information exposure.

IP-based access control restrictions in F5 BIG-IP httpd do not uniformly apply to all endpoints, allowing unauthenticated remote attackers from blocked IP addresses to access protected resources and disclose sensitive information. The vulnerability affects default configurations where network-based access policies are expected to enforce restrictions across the entire application stack, but certain endpoints bypass these controls. A vendor patch is available.

ELECOM wireless LAN access point models WAB-BE187-M, WAB-BE72-M, WAB-BE36-M, and WAB-BE36-S fail to validate the language parameter in administrative pages, allowing remote attackers to break the admin interface for logged-in users via malicious web pages. The vulnerability requires user interaction (viewing a malicious page while authenticated to the access point) and results in denial of service of the administrative interface rather than data exposure or unauthorized access. No public exploit code has been identified at time of analysis.

ELECOM wireless LAN access point devices WRC-X1800GS-B, WRC-X3000GS2 series, WRC-X6000QS series, and related models use a hard-coded cryptographic key to encrypt configuration file backups. An attacker who obtains a backup file can decrypt and modify the configuration using the publicly known key, then trick a network administrator into restoring the malicious configuration, enabling complete compromise of network settings. This requires user interaction (administrator deploying a crafted backup) but no authentication, making it a practical attack vector for supply-chain compromise or insider threats. CVSS 6.5 (Medium) reflects the high integrity impact balanced against the requirement for administrator interaction.

csync2 compiled with C99 or later creates insecure temporary directories vulnerable to time-of-check-time-of-use (TOCTOU) attacks, allowing local authenticated users with user interaction to cause denial of service through symlink or directory manipulation. The vulnerability affects OpenSUSE Tumbleweed and requires local access with low privileges and user interaction to exploit.

Remote code execution in Pentaho Data Integration & Analytics affects all versions through vulnerable H2 database JDBC driver. Authenticated data source administrators can execute arbitrary external scripts during database connection creation, achieving complete system compromise with potential container escape (CVSS scope changed). EPSS data not provided; no CISA KEV listing identified at time of analysis. Vendor advisory indicates patches available in versions 10.2.0.7 and 11.0.0.0.

Improper privilege management in Samsung System Support Service prior to version 8.0.8.0 allows a local low-privileged attacker to invoke privileged system functions without adequate authorization checks. The CVSS 4.0 subsequent-system metrics (SC:H/SI:H/SA:H) indicate that while the vulnerable service component itself has limited direct exposure, the downstream blast radius extends to high-impact confidentiality, integrity, and availability compromise of broader system components. No public exploit has been identified and EPSS sits at 0.01% (2nd percentile), with CISA SSVC confirming no known exploitation at time of analysis.

Improper permission handling in the Routines automation feature on Samsung Mobile Devices exposes sensitive information to local attackers without requiring elevated privileges. Affected devices run Samsung One UI on Android 15 and Android 16, prior to the SMR May-2026 Release 1 security update. No public exploit has been identified at time of analysis, and CISA's SSVC framework classifies current exploitation as none, consistent with an EPSS score at the 3rd percentile.

Improper input validation in the Routines feature on Samsung Mobile Devices (Android 16) enables a physical attacker to launch privileged activities outside normal authorization boundaries. Affected devices are those running any build prior to SMR May-2026 Release 1, as confirmed by Samsung Mobile Security. No public exploit exists at time of analysis, and EPSS at 0.04% (13th percentile) reflects the negligible likelihood of widespread exploitation given the mandatory physical access requirement.

OmaCP on Samsung Android devices (Android 14, 15, and 16) exposes improperly exported application components that allow any co-resident local application to invoke privileged functions without holding the required permissions. The flaw is rooted in Android's IPC mechanism, where components lacking proper access control declarations can be targeted by crafted intents from other installed apps. No public exploit code has been identified at the time of analysis, and EPSS sits at 0.01% (3rd percentile), consistent with SSVC's exploitation status of none; risk is real but bounded by the local attack vector.

Incorrect privilege assignment in the LocationManager component of Samsung Mobile Devices (Android 14, 15, and 16) enables local attackers to access sensitive information without holding the requisite Android permissions. The flaw is addressed in Samsung's SMR May-2026 Release 1 security update. No public exploit code exists, EPSS stands at the 3rd percentile, and CISA has not listed this in the KEV catalog, placing it in a low-urgency category despite the broad Samsung device footprint.

The Broadstreet WordPress plugin versions up to 1.53.1 exposes sensitive business information through an unauthenticated AJAX endpoint (get_sponsored_meta), allowing attackers to extract password-protected and private business details. Despite the CVSS vector indicating PR:N, the vulnerability requires subscriber-level or higher WordPress access, making authenticated users the primary attack vector. The exposure is limited to confidentiality impact with no integrity or availability compromise.

Improper enforcement of the LFENCE serialization property may allow an attacker to bypass speculation barriers and potentially disclose sensitive information, potentially resulting in loss of. Rated medium severity (CVSS 5.7). No vendor patch available.

Missing hardware lock bit protection on NBIO registers in AMD EPYC server processors (7003/8004/9004/9005 series, including embedded variants) allows a local admin-privileged attacker to reconfigure MMIO routing, undermining SEV-SNP guest integrity guarantees. The vulnerability is significant in confidential computing contexts where SEV-SNP is deployed to protect tenant workloads from potentially malicious host operators - precisely the trust boundary AMD's hardware security feature is designed to enforce. No public exploit exists and EPSS (0.01%, 2nd percentile) combined with SSVC exploitation status of 'none' indicate negligible near-term mass exploitation risk; however, targeted abuse by privileged insiders or compromised hypervisor hosts is the realistic concern.

Kernel and cross-guest memory disclosure in VMware ESXi 8.x/9.x hosts running the AMD-Pensando 'ionic' cloud DPU driver allows an attacker operating from an unprivileged guest VM to read hypervisor kernel memory or the memory of co-located guest VMs, breaking tenant isolation. The flaw (CWE-822, untrusted pointer dereference) was reported by AMD and can result in loss of confidentiality or availability. There is no public exploit identified at time of analysis, and EPSS is negligible at 0.01%, with CISA SSVC recording no observed exploitation.

Use-after-free in MongoDB Server's Field-Level Encryption query analysis component allows authenticated remote attackers with control over FLE query structure to cause information disclosure and denial of service. The vulnerability affects mongocryptd and crypt_shared in versions 7.0 prior to 7.0.34, 8.0 prior to 8.0.23, 8.2 prior to 8.2.9, and 8.3 prior to 8.3.2. No public exploit code identified at time of analysis.

MongoDB Server fails to fully redact user data in local server log messages when schema validation is enabled and an update or insert operation violates the collection schema, allowing authenticated administrators to access sensitive information through log inspection. This information disclosure affects MongoDB Server 7.0 prior to 7.0.34, 8.0 prior to 8.0.23, 8.2 prior to 8.2.9, and 8.3 prior to 8.3.2. The vulnerability requires high-privilege administrative access and has a low CVSS score of 2.7, indicating limited real-world impact despite confirmed patch availability.

Arbitrary file read on Garmin WDU v1 1.4.6 and v2 5.0 allows remote unauthenticated attackers to retrieve sensitive files from the device filesystem via symlink injection in uploaded graphics packages. The locally-served web server follows symlinks without filesystem restriction, enabling information disclosure. EPSS score of 0.02% (5th percentile) indicates low widespread exploitation probability. No public exploit or CISA KEV listing identified at time of analysis.

SSO login flow in Warpgate bastion host prior to 0.23.3 omits validation of the OAuth state parameter, enabling a cross-site request forgery attack that forces a victim to authenticate into the attacker's account instead of their own. Once logged in as the attacker, the victim may unknowingly write sensitive data to attacker-controlled SSH targets or interact with attacker-configured HTTP targets, resulting in high integrity impact with scope change. No confirmed active exploitation (not in CISA KEV), though SSVC assessment indicates a proof-of-concept exists; EPSS at 0.01% reflects low real-world exploitation probability at time of analysis.

Arbitrary code execution and information disclosure in the Ashlar-Vellum CAD/3D-modeling suite (Cobalt, Xenon, Argon, Lithium, and Cobalt Share, versions 12.6.1204.216 and prior) can be triggered when the application parses a maliciously crafted VC6 file. The flaw is an out-of-bounds read (CWE-125) reached through the file-import path, giving an attacker who convinces a user to open a booby-trapped file the ability to leak memory or run code in the context of the victim. There is no public exploit identified at time of analysis and EPSS is negligible (0.01%), consistent with the local, user-interaction-dependent attack model reported by CISA ICS-CERT.

Arbitrary code execution and information disclosure in Ashlar-Vellum's CAD suite (Cobalt, Xenon, Argon, Lithium, and Cobalt Share, versions 12.6.1204.216 and prior) is possible when the application parses a specially crafted VC6 file, driven by an out-of-bounds read (CWE-125). An attacker who convinces a user to open a malicious VC6 file can read out-of-bounds memory to leak data or steer execution toward running arbitrary code in the context of the user. There is no public exploit identified at time of analysis, the EPSS score is negligible (0.01%), and CISA's SSVC framework rates exploitation as 'none' but technical impact as 'total'.

Stack-based out-of-bounds read in NanaZip's ZealFS v1 filesystem image parser affects versions 5.0.1252.0 through before 6.0.1698.0, allowing an attacker who tricks a user into opening a crafted archive to leak stack memory and potentially crash the application. The root cause is an attacker-controlled BitmapSize header field that drives an unbounded loop reading past the end of a stack-allocated ZEALFS_V1_HEADER structure, yielding limited confidentiality and availability impact. No active exploitation has been confirmed - EPSS is 0.01% (2nd percentile), SSVC rates exploitation as none, and no KEV listing or public exploit code exists at time of analysis.

Arbitrary file read and write on the Nomad client host is achievable through a symlink attack in HashiCorp Nomad's exec2 task driver prior to version 0.1.2. An attacker operating as the Nomad process user with high local privileges can exploit CWE-59 link-following behavior to redirect file operations outside the intended task sandbox, reading or writing arbitrary files on the underlying host. No public exploit code has been identified and CISA has not added this to the Known Exploited Vulnerabilities catalog, but the integrity impact is significant given the scope change to the host system.

Arbitrary file read and write on Nomad client hosts is possible through a symlink attack (CWE-59) affecting HashiCorp Nomad and Nomad Enterprise versions 0.9.0 through 1.x and 2.x prior to fixed releases. A local attacker operating with high privileges - such as a Nomad operator - can plant a malicious symlink that Nomad follows without proper validation, causing the Nomad process to read or write files outside its intended scope on the host filesystem, constituting a scope change per the CVSS vector (S:C). No public exploit exists and CISA has not added this to the KEV catalog; EPSS at 0.01% (1st percentile) and SSVC exploitation status of 'none' confirm this as low active-exploitation risk despite the scope change elevating CVSS to 6.0.

XML entity expansion (CWE-776) in the DHCP XML handling component of HPE Aruba AOS Instant 8.x.x.x enables an unauthenticated remote attacker to cause denial-of-service via excessive resource consumption on affected Access Points. Exploitation requires both high attack complexity and user interaction (CVSS AV:N/AC:H/PR:N/UI:R), restricting the realistic attack surface to scenarios where a connected wireless client triggers the vulnerable DHCP parsing path. No active exploitation has been identified - EPSS stands at 0.07% (21st percentile), SSVC confirms exploitation status of none, and the vulnerability is absent from CISA KEV.

Out-of-bounds read in Adobe Illustrator 29.8.6, 30.3 and earlier exposes sensitive process memory to disclosure when a victim opens a specially crafted file. The CVSS vector confirms a local-execution context with high confidentiality impact and no integrity or availability impact, consistent with a file-parser memory leak rather than code execution. No public exploit code or active exploitation has been identified at time of analysis, and EPSS at 0.02% (6th percentile) reflects minimal observed exploitation pressure.

Sensitive user data disclosure in Apple iOS and iPadOS before 18.7.3 and before 26.2 stems from an inconsistent user-interface state that allows a malicious or crafted app to access data it should not. The flaw, reported by Apple's own product security team and fixed across the iOS 18.7.x and 26.x branches, carries a CVSS of 7.5 (high confidentiality impact) but an EPSS of just 0.03%, and there is no public exploit identified at time of analysis. CISA SSVC rates exploitation as none, indicating a real but low-urgency issue rather than an actively targeted one.

Local privilege escalation in Azure Monitor Agent Metrics Extension (versions below 1.42.0) allows an authenticated low-privileged local attacker to elevate privileges by exploiting an untrusted search path condition (CWE-426). The agent process resolves executable or library paths from an attacker-controllable location, enabling substitution of malicious code that executes in a higher-privileged security context (S:C). No public exploit code has been identified at time of analysis, and EPSS stands at 0.05% (17th percentile), indicating minimal observed exploitation pressure.

Absolute path traversal in Microsoft .NET Core (Windows runtime, versions 8.0 through 10.0) enables a network-accessible attacker to write arbitrary files and directories to restricted locations on a vulnerable system by sending a specially crafted file. Exploitation is constrained by limited attacker control over the write destination, and SSVC assessment rates exploitation as 'none' with non-automatable impact, consistent with the low EPSS of 0.08%. A publicly available proof-of-concept exists on GitHub Gist, but no confirmed active exploitation appears in CISA KEV; patched runtime versions are available for all affected .NET branches.

Microsoft Teams for Android exposes internal files or directories to unauthorized local parties via CWE-552 (Files or Directories Accessible to External Parties), enabling a local attacker to perform identity spoofing. Affected versions span all releases of Teams for Android below 1.0.0.2026092103. No public exploit code has been identified and CISA has not added this to the Known Exploited Vulnerabilities catalog, though the SSVC framework rates the technical impact as total, suggesting that successfully accessed data could fully compromise the Teams session or identity context on the device.

Privilege escalation via double free in the Windows Rich Text Edit Control (RichEdit) affects an exceptionally broad range of Microsoft platforms, spanning Windows 10 versions 1607 through 22H2, all current Windows 11 branches through 26H1, and Windows Server 2012 through Server 2025. A low-privileged local attacker who can induce user interaction with crafted RTF content can corrupt heap memory to achieve full compromise of confidentiality, integrity, and availability - consistent with the SSVC 'total technical impact' rating. No public exploit code exists and no active exploitation has been identified (not in CISA KEV), and the EPSS score of 0.05% at the 16th percentile reflects a minimal current exploitation signal; however, the sheer scale of the affected Windows install base keeps this relevant for enterprise patch cadence. A notable discrepancy exists: the CVE tags classify this as 'Information Disclosure,' which conflicts with Microsoft's own description of local privilege elevation - the CVSS C:H/I:H/A:H impact metrics and SSVC total-impact rating support the privilege escalation characterization, suggesting the tag is erroneous.

External control of file name or path (CWE-73) in Microsoft Office Word enables network-based information disclosure when a victim opens a specially crafted document, affecting Word 2016, Office 2019, LTSC 2021/2024, and Microsoft 365 Apps for Enterprise. An unauthenticated remote attacker can exploit this to coerce Word into resolving an attacker-controlled file path - likely a UNC path - causing an outbound network request that leaks sensitive data such as NTLM credentials. No active exploitation is confirmed (not listed in CISA KEV), and the EPSS score of 0.05% at the 14th percentile indicates negligible real-world exploitation activity at time of analysis.

Information disclosure in Microsoft Power Automate for Desktop (all versions from 1.0.0.0 up to but not including 2.67) enables an authenticated, low-privileged attacker operating over the network to access sensitive data they are not authorized to view. The flaw, rooted in CWE-200 improper exposure of sensitive information, carries a CVSS 6.5 score reflecting high confidentiality impact with no integrity or availability consequence. No active exploitation has been confirmed - EPSS stands at 0.11% and CISA KEV listing is absent - but patch availability from Microsoft makes remediation straightforward.

Local file disclosure in Microsoft Word exposes sensitive files and directories to external parties when a victim opens a specially crafted document. Affected products span Microsoft Word 2016, Office 2019, Office LTSC 2021, Office LTSC 2024, and Microsoft 365 Apps for Enterprise across version ranges confirmed by NVD CPE data. No public exploit code has been identified and CISA SSVC rates exploitation as none with only partial technical impact; a vendor-released patch is available via the Microsoft Security Response Center.

Out-of-bounds read in the Windows Telnet Client exposes memory contents and minor availability degradation to network-based attackers who control a Telnet server endpoint. Affected systems span Windows 10 (1607 through 22H2), Windows 11 (22H3 through 26H1), and Windows Server 2012 through 2025, all requiring user interaction to initiate a Telnet session against a malicious server. No public exploit identified at time of analysis and no active exploitation confirmed (not in CISA KEV); EPSS stands at 0.05% (16th percentile), reinforcing that real-world exploitation risk is low despite the broad platform footprint.

Out-of-bounds read in the Windows Desktop Window Manager (DWM) Core Library exposes memory contents to a locally authenticated attacker on Windows 11 (versions 24H2, 25H2, 26H1) and Windows Server 2025. The flaw carries a CVSS score of 5.5 with high confidentiality impact but no integrity or availability consequence, making it a pure information disclosure primitive. No public exploit code has been identified and CISA SSVC marks exploitation status as none, with EPSS at 0.04% (11th percentile), indicating very low near-term exploitation probability.

Double free vulnerability in Windows Rich Text Edit component allows local authenticated attackers to escalate privileges on Windows 10 and Windows 11 systems through a specially crafted interaction. The flaw requires local access with standard user privileges and user interaction, but enables full system compromise including code execution and privilege elevation. Microsoft has released a vendor patch to address this issue.

Improper export of Android application components in Fortinet FortiToken Android 5.2, 6.1, and 6.2 allows local authenticated attackers to gain unauthorized access to sensitive information via exposed application components that lack proper access control. The vulnerability has a CVSS score of 5.0 with local attack vector and requires low privileges, enabling information disclosure without user interaction. No public exploit code has been identified, and the vulnerability is not listed in active exploitation databases at the time of analysis.

Fortinet FortiClient Windows versions 7.2 (all) and 7.4.0 through 7.4.2 contain a hard-coded cryptographic key vulnerability that allows high-privileged local attackers to disclose sensitive information. The vulnerability requires local access and administrator-level privileges, limiting its real-world exploitation scope to threats already present on compromised systems or malicious insiders. No public exploit code or active exploitation has been confirmed at the time of analysis.

Improper initialization in Intel UEFI firmware (Ring 0 / Bare Metal OS layer) enables information disclosure on some Intel platforms. A system software-level adversary already holding high privileges can exploit this flaw under high-complexity, prerequisite-dependent conditions to expose sensitive memory contents, with confidentiality impact rated high but no integrity or availability effect. No public exploit exists and EPSS sits at 0.01% (2nd percentile), and SSVC classifies exploitation status as none - signals that are mutually consistent and indicate this is a low-urgency finding for most organizations.

Transient execution information disclosure in Intel processors allows unprivileged guest VM code to leak sensitive data from co-resident workloads by exploiting shared microarchitectural predictor state during VMX non-root operation. Confidentiality impact is rated high at both the vulnerable-system and subsequent-system level, indicating potential cross-VM or host-to-guest data leakage consistent with Spectre-class side-channel behavior. No public exploit has been identified at time of analysis, EPSS sits at the 3rd percentile, and CISA SSVC rates exploitation as none - but the cross-boundary confidentiality impact makes this relevant to any multi-tenant virtualized environment running affected Intel silicon.

Timing side-channel in Apache Tomcat's AJP secret comparison exposes the shared AJP connector secret to remote, unauthenticated attackers capable of making precise network timing measurements. The vulnerability, tracked as CWE-208 (Observable Timing Discrepancy), affects all major Tomcat branches from 7.0.0 through current releases prior to the fixed versions, and could allow an attacker to recover the AJP shared secret through repeated probing. No public exploit code exists at time of analysis, EPSS is 0.02%, and CISA SSVC rates exploitation as none - making real-world risk low despite the network-accessible attack vector.

protobufjs versions 7.5.5 and earlier, and 8.0.0-8.0.1 accept overlong UTF-8 byte sequences in the minimal UTF-8 decoder used by non-Node and fallback decoding paths, allowing attackers to bypass byte-level filtering and decode strings containing characters that were not present in the raw protobuf binary input. This integrity issue affects applications that rely on pre-decoding byte validation before using protobuf strings in security-sensitive contexts. Patch versions 7.5.6 and 8.0.2 are available; Node.js Buffer-backed paths are not directly affected.

Remote authenticated attackers can exploit an exposed dangerous method on the Core Server of Ivanti Endpoint Manager versions before 2024 SU6 to leak access credentials. The vulnerability requires valid authentication credentials to exploit and does not allow code execution or system modification, but compromises confidentiality by exposing sensitive authentication material that could facilitate lateral movement or account takeover.

Ivanti Secure Access Client before version 22.8R6 allows local authenticated users to read or modify sensitive log data through write access to a shared memory section due to incorrect permission assignments on a critical resource. With a CVSS score of 4.4 and a local attack vector requiring authentication, this vulnerability poses a moderate risk to users whose systems are accessed by multiple authenticated accounts. No active exploitation has been publicly confirmed, but the simplicity of the attack (local, low complexity) makes this a practical concern for multi-user systems.

Credential leakage in LWP::UserAgent before 6.83 (Perl) exposes Authorization and Proxy-Authorization headers to attacker-controlled redirect targets across cross-origin 3xx redirects. The library's redirect handler stripped only Host and Cookie on follow-up requests, leaving credential headers intact even when the redirect crossed a scheme, host, or port boundary. Authenticated Perl HTTP clients - including server-side applications, crawlers, API integrators, and automation tooling - are affected whenever caller-supplied credentials are passed to a UserAgent instance that can be redirected. No public exploit has been independently confirmed beyond the proof-of-concept submitted with the vulnerability report, and CISA KEV does not list this CVE; however, the exploitation pattern is straightforward and mirrors a well-documented class of credential-leakage flaws in HTTP client libraries.

Man-in-the-middle attackers positioned between OX Dovecot Pro and clients can forge SCRAM TLS channel binding via specially crafted base64 exchanges, allowing eavesdropping on encrypted communications. The attack requires network-level access and knowledge of channel binding mechanics but yields complete confidentiality compromise. No public exploit code is known, and patched versions are available from Open-Xchange.

Information disclosure vulnerability in Firefox's JavaScript Engine allows remote unauthenticated attackers to leak sensitive memory contents over the network without user interaction. The vulnerability affects Firefox versions prior to 150.0.3 and has a low EPSS score (0.02%) despite the network-based attack vector, suggesting limited real-world exploitation pressure despite the modest CVSS score of 5.3.

WP EasyPay plugin through version 4.3.0 exposes sensitive information in sent data, allowing unauthenticated remote attackers to retrieve embedded data without user interaction. The vulnerability stems from improper handling of sensitive data during transmission, classified as an information disclosure issue with a CVSS score of 5.3 (network-accessible, low complexity). No active exploitation has been confirmed in CISA KEV at the time of analysis.

Hikvision Hik-Connect App fails to enforce strict directory access permissions, allowing other locally-installed applications to read sensitive information stored by the app on the device filesystem. The vulnerability requires local access and high attack complexity but poses information disclosure risk to users with multiple apps on the same device. No public exploit or active exploitation has been identified; CVSS score of 2.9 reflects the local-only attack vector and limited confidentiality impact.

Authenticated remote attackers can read arbitrary files from the operating system filesystem with root privileges on affected RUGGEDCOM ROX devices due to improper input validation in the JSON-RPC web server interface. All versions of ROX MX5000, MX5000RE, RX1400, RX1500, RX1501, RX1510, RX1511, RX1512, RX1524, RX1536, and RX5000 below V2.17.1 are affected. No public exploit code has been identified at time of analysis.

A vulnerability has been identified in SIPROTEC 5 6MD84 (CP300) (All versions < V11.0), SIPROTEC 5 6MD85 (CP200) (All versions), SIPROTEC 5 6MD85 (CP300) (All versions >= V7.80 < V11.0), SIPROTEC 5. Rated medium severity (CVSS 6.9), this vulnerability is remotely exploitable, no authentication required, low attack complexity. No vendor patch available.

Slek Gateway for WooCommerce plugin version 1.0 exposes merchant API credentials (slek_key and slek_secret) to unauthenticated attackers through client-side HTML forms and plaintext GET parameters. An attacker who places an order on an affected WooCommerce store can extract the merchant's secret credentials by inspecting the HTML source or using browser developer tools on the order-pay page before JavaScript auto-submission occurs, compromising the merchant's Slek payment processing account.

Cross-origin source code exposure in webpack-dev-server up to 5.2.3 allows attackers controlling a malicious website to steal bundled application source code when a developer runs the dev server over non-trustworthy HTTP origins. The vulnerability exploits the omission of Sec-Fetch-Mode and Sec-Fetch-Site headers on non-HTTPS connections, enabling script injection and cross-origin code exfiltration. Chromium-based browsers Chrome 142+ are exempt due to local network access restrictions. CVSS 5.3 (AC:H due to user requirement to visit attacker site; High confidentiality impact). Fix: upgrade to webpack-dev-server 5.2.4 or later.