Missing authorization in the REDAXO backup addon's database export (core versions up to and including 5.21.4) lets a non-admin user who has been granted the backup[export] role permission export any table in the database, including rex_user, rex_user_session, rex_user_passkey and tables belonging to other applications sharing the same database. Because rex_user_session stores session IDs and stay-logged-in keys in plaintext, such a user can impersonate any account with an active session or persistent login - including administrators - and rex_user exposes password hashes for offline cracking, escalating the attacker from a limited role to full admin control. Exploitation is gated on an administrator explicitly delegating backup[export] to a non-admin account (by default only admins reach the backup addon), and full session-based takeover additionally requires a victim admin to have a live or stay-logged-in session; no public exploit code was identified at time of analysis, and the vendor fixed the issue in 5.21.5.
Octopus Server writes user-invitation codes into audit event logs, so an authenticated user who can read audit events in a space may retrieve the code for an unclaimed, space-linked invitation and redeem it. The resulting privilege gain depends entirely on the permissions assigned to the invited account, and exploitation also requires that an unredeemed space-linked invitation already exists; the independent assessment therefore treats this as a moderate-priority authenticated privilege-escalation issue rather than a mass-exploitation threat. Vendor-released fixes are 2026.2.13438 and 2026.3.15829, and no public exploit is identified in the supplied data.
Improper certificate validation in Moxa MRC Client 1.0.0 through 4.2.0 allows an unauthenticated attacker who already holds an on-path position between the client and its MRC server to impersonate the trusted server and conduct a man-in-the-middle attack, compromising confidentiality and integrity but not availability. The attack requires that network position and passive user interaction, so it is not a mass-exploitation profile; no public exploit was identified at time of analysis and EPSS is 0.17% (7th percentile). The priority signal is this_week, consistent with a moderate-priority issue rather than an active-exploitation emergency.
A remote, unauthenticated denial of service exists in Asterisk releases after 20.x: the Q.931 party-number parser in the optional chan_ooh323 addon channel driver omits a minimum-length check, so malformed Q.931 elements in an OOH323 request crash the Asterisk process (CVSS 3.1 7.5, availability-only). Exposure is narrow by design - the addon is not built by default, must be explicitly compiled and installed, and its signalling port must be reachable by the attacker - but any deployment meeting all three conditions is fully exposed to an unauthenticated network attacker. Fixed builds are 21.12.3, 22.10.1, 23.4.1 and Certified Asterisk 22.8-cert3; no public exploit identified at time of analysis, and the issue is not in CISA KEV.
Password-protected link shares in the 3D Viewer for Nextcloud app (versions 1.7.9 through 3.3.1) can be read without any password or session by anyone who knows the share token, via GET /ocs/v2.php/apps/threedviewer/public/file/{token}/{fileId}. The app's public controller was not a Nextcloud PublicShareController, so the framework's PublicShareMiddleware returned early and never enforced the share password, expiry, or share type; requests returned 200 with the full file body for shares whose normal Nextcloud share page still prompted for a password. Exposure is limited to files in the 3D formats the app handles (STL, GLB/glTF, OBJ, STEP, IFC and similar); non-3D files in the same share are unaffected and expired shares return 500 rather than leaking content. No public exploit code was identified at time of analysis; the vendor fixed the issue in 3.3.2.
Supermicro's X12SPI BMC firmware (01.07.20) does not enforce the integrity and encryption requirements it is supposed to apply to IPMI commands, so an attacker who can reach the management interface can craft packets that execute arbitrary commands inside an already-active IPMI session. The CVSS vector is unauthenticated (PR:N), but exploitation is conditional: the BMC must be network-reachable and a live session must already exist for the attacker to act against, which is why the assessed severity combines high confidentiality, integrity and availability impact with high attack complexity. There is no KEV listing, EPSS is 0.34% (27th percentile) and no public exploit code was identified at time of analysis; the vuln.today decision label is this_week, a genuine but non-emergency priority.
Unauthenticated requests against Nextcloud Server's federated share endpoint (Server 22.0.0 through 34.0.0 per EUVD) can be abused to insert arbitrary rows into the oc_share_external table, and because that table has no automatic cleanup the injected rows persist indefinitely. Three distinct denial-of-service conditions follow from this write primitive, all triggered from the same unauthenticated source and all re-firing on a 24-hour cycle for as long as the rows remain in the database. Severity is driven entirely by availability impact (CVSS 3.0 base 7.5, CWE-400, C:N/I:N/A:H); there is no confirmed active exploitation (not in CISA KEV) and no public exploit code identified at time of analysis, though exploitation code could not be ruled out from the HackerOne report referenced.
Crafted HTTP requests sent to the HTTPS management (GUI) interface of FortiOS 8.0.0 and FortiOS 7.6.4 through 7.6.6 can stall that interface through improper locking, denying administrative access to the device. Per the CVSS vector (AV:N/AC:L/PR:N/UI:N) the attack is remote, low-complexity, requires no authentication and no user interaction, but impact is confined to availability of the management plane (A:H; C:N/I:N) rather than traffic forwarding. Exploitation is not confirmed - SSVC records exploitation as none, EPSS is 0.28% (21st percentile) and no public exploit or PoC reference appears in the provided sources - so risk is best characterised as a moderate, exposure-dependent availability issue; the decisive factor is whether the HTTPS admin interface is reachable from untrusted networks.
Information disclosure in Hitachi HiRDB Server can be triggered by remote, unauthenticated attackers over the network, with high confidentiality impact and no integrity or availability impact (CVSS 7.5). The flaw is classified as an origin validation error (CWE-346) and affects HiRDB Server Version 10 before 10-12-01, HiRDB Server Version 9 through 09-66-22, HiRDB Server with Additional Function Version 9 through 09-66/V, HiRDB Server Version 9(32) through 09-66-22, and HiRDB Server with Additional Function Version 9(32) through 09-66-/V. EPSS is low at 0.21% (10th percentile), there is no CISA KEV confirmation, and no public exploit was identified at time of analysis; the vendor has released a patch.
Type confusion in Microsoft Edge (Chromium-based) before 149.0.4022.62 can allow a remote, unauthenticated attacker to execute code in the browser process when a victim opens or interacts with attacker-controlled web content. Exploitation is constrained by high attack complexity and mandatory user interaction, and no stated sandbox escape is involved. Microsoft has released a vendor patch; no public exploit code or confirmed active exploitation was identified in the available data, making this a prompt-patching item rather than an emergency.
Read-only ARI users on Asterisk can perform write operations through the REST-over-WebSocket feature, which allows arbitrary module path loading and, conditionally, remote code execution. Affected are Asterisk branches up to 20.20.0, 21.12.2, 22.10.0 and 23.4.0, plus Certified Asterisk up to 22.8-cert2; fixed releases are 20.20.1, 21.12.3, 22.10.1, 23.4.1 and 22.8-cert3. Exploitation is gated on three conditions holding simultaneously - the Asterisk HTTP web server explicitly enabled (it is disabled by default), the attacker able to reach that server (it binds only to 127.0.0.1 by default), and the attacker already holding valid read-only ARI credentials - which is why the assessed vector carries AC:H; no public exploit identified at time of analysis, and a vendor patch is available.
A NULL pointer dereference in Asterisk's Asterisk Manager Interface (AMI) handling of Digest Authentication over HTTP lets an unauthenticated remote attacker crash the telephony server (CVSS 3.1 7.5, AV:N/AC:L/PR:N/UI:N/A:H) by triggering a logging path in main/manager.c that dereferences a NULL address. Only deployments that have opted into all three non-default features are exposed - the HTTP web server, AMI itself, and AMI access via HTTP - and the HTTP listener binds to 127.0.0.1 by default even when enabled, so practical reachability requires the operator to have rebound it to a routable interface. No public exploit code and no confirmed active exploitation (CISA KEV) were identified at time of analysis, and the deployment preconditions mean the base score overstates real-world exposure; fixed releases exist in every maintained branch and the vendor decision is to remediate this week.
Uncontrolled resource consumption in Tanium Enforce allows remote attackers to degrade or exhaust service availability across the 2.9, 2.10, and 3.0 branches, rated CVSS 7.5 with availability-only impact (C:N/I:N/A:H). The assessed vector (AV:N/AC:L/PR:N/UI:N) describes a network-reachable, low-complexity attack requiring no privileges or user interaction, though the vendor description does not name the specific feature, protocol, or request type that triggers the exhaustion, and practical exposure is bounded by network reachability of an endpoint-management console normally deployed inside a managed enterprise perimeter. Vendor patches are available in Enforce 2.9.736, 2.10.778, and 3.0.373; no public exploit code or confirmed active exploitation was identified at time of analysis, and no EPSS or KEV data was supplied.
Denial of service in Tanium Provision allows unauthenticated remote attackers to degrade availability of affected Provision instances - 1.17 before 1.17.150, 1.23 before 1.23.161 and 1.24 before 1.24.199 - with CVSS 7.5 (AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H), so confidentiality and integrity are unaffected. Tanium has released fixed builds (1.17.150, 1.23.161 and 1.24.199); no public exploit was identified at time of analysis and the issue is not in CISA KEV. Priority is this_week: the practical risk hinges on whether the Provision service is reachable from untrusted networks or confined to a management segment.
Denial of service in Tanium Enforce (3.0 before 3.0.373, 2.10 before 2.10.778, and 2.9 before 2.9.736) lets remote attackers consume resources by getting a highly compressed payload into a decompression path, a compression bomb (CWE-409) carrying an availability-only impact and a 7.5 severity score. Fixed builds are published by the vendor, and no public exploit code has been identified at time of analysis. The precise feature or input that performs the decompression is not described in the available sources, so the practical precondition remains the affected version plus the ability to deliver a crafted payload to that path.
A stack buffer overflow in the ooTrace() trace-logging routine of Asterisk's optional chan_ooh323 H.323 channel driver lets remote unauthenticated attackers crash the PBX with a single crafted H.323 packet, producing a denial of service (CVSS 3.1 7.5, AV:N/AC:L/PR:N/UI:N, availability-only). Only installations that explicitly compiled and installed the chan_ooh323 addon - it is not part of a default build - and that expose its H.323 port to untrusted networks are affected; every other Asterisk deployment is out of scope. No public exploit code or active exploitation has been identified at time of analysis, and no EPSS score was supplied, but vendor fixes are already published for all supported branches. The vendor advisory and the CVE description frame the outcome as a process crash, not confirmed code execution, despite the CWE-787 out-of-bounds write classification.
Authentication bypass in Keepnet Labs Agentic AI / Agentic Ally lets a network-reachable, unauthenticated attacker spoof an identity and reach functionality that should require authentication, with a high confidentiality impact and no stated integrity or availability effect (CVSS 3.1 base 7.5, CWE-290). Every deployment running a build before commit 0ef61f2e is affected, and the practical exposure is bounded by how widely this niche product is deployed and internet-facing rather than by any attacker prerequisite. No confirmed active exploitation (CISA KEV) and no public exploit code were identified at time of analysis; the vendor has published a fix point at commit 0ef61f2e, and the vulnerability is prioritized as this_week.
Account enumeration in Rolantis Agentis before 4.8 lets remote unauthenticated attackers infer whether accounts exist by observing differing responses from account-handling functionality. The flaw is a CWE-204 observable response discrepancy with low confidentiality impact and no integrity or availability impact, and it is fully remediated by upgrading to 4.8. No public exploit identified at time of analysis.
Eaton easyE4 firmware versions below 2.46 ships an insufficiently strong password policy (CWE-521), so an attacker who has already obtained a stored password hash - for example by way of CVE-2026-41020 - can mount an offline password-guessing attack and potentially recover the underlying user credentials. The weakness is not exploitable on its own: the vendor's CVSS 3.1 vector (AV:N/AC:H/PR:N/UI:R/S:U/C:H/I:H/A:H, 7.5) reflects the hash-disclosure prerequisite and the dependence on how strong the configured password actually is, and strong, high-entropy passwords resist offline cracking even under the weak policy. No public exploit code has been identified at time of analysis, and the exposure is limited to easyE4 devices that have not been updated to firmware 2.46 or later.
Eaton easyE4 controllers before version 2.46 reuse a static nonce during authentication, allowing an attacker who can capture a legitimate authentication exchange to replay it and gain unauthorized access to the device. The CVSS 3.1 score is 7.5 (AV:N/AC:H/PR:N/UI:R), reflecting that exploitation depends on a real login occurring and being observed on the control-network path; no public exploit code or CISA KEV entry was identified at time of analysis. Eaton has released a fix in easyE4 2.46.
Unauthenticated remote attackers can establish an RMCP+ session on Supermicro MBD-X12SPI baseboard management controllers running BMC firmware 1.07.20 by sending a request immediately followed by a specially crafted RAKP message, abusing incorrect ordering in the RAKP+ authentication handshake; a session established this way could then be used to execute commands. The flaw is rated CVSS 7.4 (AV:N/AC:H/PR:N/UI:N, high confidentiality and integrity impact, no availability impact) and requires network reachability to the BMC's RMCP+/IPMI 2.0 interface, typically UDP 623. EPSS is 0.45% (39th percentile), the CVE is not listed in CISA KEV and no public exploit has been identified at time of analysis; this is a genuine but second-tier priority rather than a drop-everything emergency.
OS command injection in the mcp-ffmpeg server (versions through 1.0.5) lets an attacker who can influence the videoPath or outputDir arguments of the resize-video, extract-audio, or get-video-info tools break out of the child_process.exec command string and run arbitrary shell commands as the user running the MCP server. The breakout characters ($(), backticks, or a closing double quote) must be carried inside a path or an already-existing file/directory name that the tool forwards into the exec string, which is why the CVSS 4.0 vector is scored 7.3 High with AV:L, AC:H, AT:P and PR:L rather than as a trivial remote unauthenticated RCE. No CISA KEV listing and no public exploit code were identified at time of analysis; the priority label is 'this week', and the practical blast radius is bounded by the privileges of the MCP server account.
OS command injection in the Supermicro BMC firmware for the MBD-X12STW-F motherboard (version 01.02.08) lets an attacker with high privileges on the management interface inject a command payload into a BMC function, with the payload executing when the related service is restarted. Because the CVSS vector specifies PR:H, exploitation requires pre-existing authenticated high-privilege access to the BMC rather than being reachable by an anonymous remote user, and execution is gated on a subsequent service restart rather than occurring immediately. No confirmed public exploit code and no CISA KEV entry are present; EPSS is 0.88% (59th percentile), and the vendor priority is this_week, making this a high-severity but access-gated issue rather than a drop-everything emergency.
Local privilege escalation in TeamViewer Full Client and Host on Windows allows an authenticated low-privileged user to execute arbitrary commands as SYSTEM by placing a crafted installer options file in the Advanced Asset Management installer staging directory, which is processed during service activation. All Full Client and Host builds prior to 15.82 are affected, along with the legacy Windows 7/8 branch prior to 15.64.10. A vendor patch is available, and no public exploit has been identified at time of analysis.
Local arbitrary code execution in Adobe Creative Cloud Desktop Application versions up to and including 6.10.0.252.41 is enabled by an uncontrolled search path element (CWE-427). A low-privileged attacker can exploit the application's component-loading behavior, but the victim must open a malicious file, and the resulting code runs only in the current user's context. The flaw is rated CVSS 7.3, no public exploit has been identified at time of analysis, and Adobe has published advisory APSB26-135.
Local privilege escalation in Veeam Agent for Linux 13.1 allows a local low-privileged user to gain elevated privileges on hosts whose agent installation is managed by a Veeam Backup & Replication instance; standalone/unmanaged deployments are not affected. The flaw is an untrusted search path (CWE-426), the vendor has released a fix in 13.1.2, and no public exploit identified at time of analysis. This is a genuine but moderate-priority issue rather than an emergency.
Local privilege escalation in TeamViewer Remote Management Patch Management for Windows allows an authenticated low-privileged user to execute arbitrary code with elevated privileges by replacing the service executable in an insufficiently protected temporary directory during Patch Management activation. The flaw affects versions prior to 26.9.1 and the legacy v14.7 client branch prior to 24.12.5. Exploitation requires a local foothold and an activation event, and no CISA KEV entry or public exploit code is referenced in the supplied intelligence.
DLL search-order/planting in the TeamViewer Remote Management Monitoring installer for Windows before 26.10.1 lets a local authenticated low-privileged user plant a malicious DLL that is loaded with elevated privileges when a higher-privileged user runs the installer, yielding arbitrary code execution. Exploitation requires an existing foothold on the host plus user interaction, and no public exploit is identified at time of analysis. Vendor patch 26.10.1 resolves the condition; the assessed priority is this_week, a genuine but not emergency-grade issue.
Path traversal in Dell PowerProtect Data Domain Virtual Edition allows a remote attacker who already holds administrative-level privileges to execute commands on the appliance. Affected releases are Feature Release 7.7.1.0 through 8.7, LTS2024 7.13.1.0 through 7.13.1.70, LTS2025 8.3.1.0 through 8.3.1.30, and LTS2026 8.6.1.0 through 8.6.1.10; the PR:H CVSS vector means unauthenticated and low-privileged users cannot reach the flaw. No public exploit was identified at time of analysis, and Dell has released fixed builds.
Missing OAuth token audience validation in AWS API MCP Server 1.3.0 through 1.5.x lets a local caller reach the server using a bearer token that the identity provider issued to a different application in the same tenant, and then run AWS operations with the AWS credentials configured on the host. The flaw is conditional: it only applies to deployments using the non-default streamable-http transport with OAuth authentication enabled against an operator-nominated identity provider, while the default stdio transport is unaffected. Resulting access is bounded by the permissions of the host's AWS credentials; no public exploit code has been identified at time of analysis, and version 1.6.0 fixes the issue and additionally requires operators to configure an expected token audience or client ID.
WatchGuard Fireware OS firewall deployments below 12.5.22, below 12.12.4, and 2025.1 before 2026.4 allow a remote, authenticated attacker with at least read-only Device Monitor privileges and management CLI access to submit a crafted diagnostic command that creates, truncates, or overwrites arbitrary files and modifies kernel parameters with root privileges. The impact is integrity and availability loss rather than confidentiality exposure, and written file content is limited to numeric values; no public exploit was identified at time of analysis.
Authenticated IPMI administrators on Supermicro MBD-X12SPI BMC firmware 01.07.20 can read and write arbitrary physical memory, leading to code execution on the BMC operating system. The issue is rated CVSS 7.2 (AV:N/AC:L/PR:H/UI:N/S:U/C:H/I:H/A:H) and classified CWE-287; exploitation requires valid IPMI Administrator credentials and network access to the BMC management interface. No public exploit code has been identified at time of analysis, EPSS is 0.32% (24th percentile), and the vuln.today priority is this_week, so this is a moderate, authentication-gated risk rather than an emergency.
Protection-mechanism bypass in Dell PowerProtect Data Domain Virtual Edition allows a remote attacker who already holds a high-privileged (administrative) account to subvert the appliance's data-protection controls. Affected builds are Feature Release 7.7.1.0 through 8.7, LTS2024 7.13.1.0 through 7.13.1.70, LTS2025 8.3.1.0 through 8.3.1.30 and LTS2026 8.6.1.0 through 8.6.1.10; the flaw is scored CVSS 7.2 (AV:N/AC:L/PR:H/UI:N/S:U/C:H/I:H/A:H) with the PR:H metric confirming the administrative precondition, which sharply limits the realistic attacker set to insiders or adversaries who have already compromised privileged credentials. No public exploit code and no CISA KEV entry are associated with this CVE at time of analysis, and vendor patches are available, so this is a scheduled-patching item rather than an emergency.
Stored cross-site scripting in the Analytics Insights - Google Analytics Dashboard for WordPress plugin allows script injection through the Page Title in versions up to and including 6.3.12, executing when a user later accesses the affected page. The supplied CVSS 3.1 vector is 7.2 (AV:N/AC:L/PR:N/UI:N/S:C/C:L/I:L/A:N), and EPSS is 0.24% at the 15th percentile, with no CISA KEV entry and no public exploit identified at time of analysis. The vector indicates unauthenticated, no-interaction exploitation, but the sources do not describe the exact endpoint or input path by which an unauthenticated attacker sets the Page Title.
Stored cross-site scripting in the Advanced Views - Display Framework - Gutenberg/Elementor/Shortcode plugin (acf-views) for WordPress, affecting all versions up to and including 3.9.6, allows unauthenticated visitors to plant script payloads through the public 'comment' parameter that later execute in the browser of anyone loading a page where an avf-layout block renders that comment's content. The decisive weakness is that the plugin's get_comment() lookup applies no approval-status filter, so a pending or unapproved comment is enough to trigger the payload - ordinary comment moderation does not mitigate the attack. Severity is rated CVSS 7.2 (AV:N/AC:L/PR:N/UI:N/S:C/C:L/I:L/A:N, CWE-79) with a this_week priority label; no public exploit code and no CISA KEV entry were identified at time of analysis.
An out-of-bounds read in WatchGuard Fireware OS's FireCluster synchronization processing lets an attacker who already sits on the cluster interconnect - a compromised cluster peer or an on-path attacker on the sync link - repeatedly crash the cluster synchronization process with crafted traffic, producing a sustained denial of service. Only Fireboxes deployed in a FireCluster high-availability configuration are exposed; stand-alone Fireboxes are not reachable this way, and the CVSS 4.0 base score of 7.2 (HIGH) is driven almost entirely by high availability impact (VA:H) with only limited confidentiality/integrity effect (VC:L/VI:L) from the out-of-bounds read itself. Fixed builds are available from the vendor (12.12.4, 12.5.22 on the 12.0 LTS branch, and 2026.4 on the 2025.1 branch); no public exploit code has been identified at time of analysis and the flaw is not listed as actively exploited.
Arbitrary file read and deletion in GRR (JeromeDevome) up to and including 4.4.2 lets an authenticated administrator abuse the database restoration feature to point at files outside the intended restore location, disclosing sensitive data or destroying files and causing denial of service. The flaw sits in the admin-only restore controller, so exploitation is gated behind a high-privilege account, and the files actually reachable or deletable depend on the OS permissions of the web server process, with impact amplified when the server runs elevated. The vendor fix is version 4.5.0; no public exploit code was identified at time of analysis.
The CTFd Terraform Provider (ctfer-io/terraform-provider-ctfd) writes the ctfd_team.password attribute into Terraform state as a non-sensitive value in versions 0.1.5 through 2.8.1, so anyone who can read that state file obtains the CTFd team password directly, without needing the state encryption password or any further decryption step. The exposure is access-gated rather than anonymous: the attacker must already be able to read the state (a shared or compromised host, a state file committed to version control or leaked through CI logs, or an exposed remote backend), which is reflected in the assessed vector CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:N and a confidentiality-only impact limited to team passwords. The vendor addressed it in v2.8.2 by marking the schema attribute Sensitive, states that no workaround exists, and no public exploit code has been identified at time of analysis.
Cross-site scripting in the CKEditor 5 engine (@ckeditor/ckeditor5-engine) allows crafted markup to execute attacker-controlled JavaScript when the editor processes it, but only on installations where General HTML Support is enabled and explicitly configured to permit inserting object content. Exposed versions are 29.0.0 through 47.7.3 and 48.0.0 through 48.5.0; default builds and deployments without that non-default GHS object allowlist are not affected. Per the vendor and the assessed vector (AV:N/AC:H/PR:L/UI:N/S:U/C:N/I:L/A:N), the injected script runs in a browser context isolated from the embedding application's origin, so the vendor models no confidentiality impact and only low integrity impact; no public exploit was identified at the time of analysis and the issue is tracked as monitor priority.
SQL injection in Asterisk's Channel Event Log database backends cel_pgsql and cel_tds allows attacker-influenced data placed in the CEL eventtype column to be executed as arbitrary SQL against the connected PostgreSQL or TDS/MSSQL database. Exposure is limited to deployments that load and configure one of those CEL backends, invoke the CELGenUserEvent dialplan application, and pass it unvalidated data that a caller or other untrusted source can influence; default installations without these backends and dialplans are not affected. Fixed releases are Asterisk 20.20.1, 21.12.3, 22.10.1, 23.4.1 and Certified Asterisk 20.7-cert11 / 22.8-cert3; no public exploit was identified at time of analysis.
Stored cross-site scripting in the Autodesk Fusion desktop application allows a maliciously crafted HTML payload embedded in the description field of an electronic library (.lbr) file to execute when that library is imported and rendered in the Library Manager. Successful exploitation can let an attacker read local files or execute arbitrary code in the context of the Fusion process, so the impact extends beyond a typical cosmetic script injection even though the attack is not remotely triggerable. Affected builds are Fusion 2705.1.30 up to, but not including, 2706.1.14; the vendor has released a fix, no public exploit has been identified at time of analysis, and the current priority decision places remediation in the this_week window.
Reflected cross-site scripting in the Autodesk Fusion desktop application (builds 2705.1.30 up to, but not including, 2706.1.14) allows an attacker to run HTML/script in the context of the Fusion process when a user invokes a crafted Fusion protocol URL whose documentName parameter carries the payload. Successful exploitation can let the attacker read local files or execute arbitrary code as the current user, so the integrity and confidentiality impact is high, though the attack requires the victim to click or open the malicious link and cannot trigger autonomously (AV:L, UI:R, PR:N). No public exploit code has been identified at time of analysis and there is no CISA KEV listing; Autodesk has released a fixed build and the vendor advisory is available.
Stored cross-site scripting in Autodesk Fusion 2705.1.30 through versions before 2706.1.14 allows an attacker who delivers a design with a crafted name to run HTML/JavaScript when another user opens and views that design, potentially reading local files or executing code in the context of the Fusion process. This is a moderate, conditional risk rather than an emergency: exploitation requires user interaction (UI:R) on a local vector (AV:L) with no privileges required (PR:N), and no public exploit was identified at time of analysis. Updating to 2706.1.14 fully addresses the issue per the assessment.
Local privilege escalation and arbitrary file overwrite in WatchGuard Fireware OS let an attacker who already holds code execution as a low-privileged system process write files with root privileges and tamper with firewall IP set data. The attacker plants a crafted filesystem object such as a link in a location the firewall daemon processes, then triggers a firewall IP-set list update so the root-privileged daemon follows it; the written content is only partially constrained and the list-update precondition must be satisfied. No public exploit code was identified at time of analysis and the flaw is not listed in CISA KEV, with fixed builds available from the vendor.
Cross-site request forgery in the web/management interface of Ubiquiti airMAX M devices running firmware below 6.3.26 allows an attacker to make unauthorized configuration changes and push the device into a denial-of-service condition by luring an already-authenticated administrator to an attacker-controlled page. No credentials for the device are required (PR:N), but the forged request only takes effect against the victim's live session, so the required user interaction (UI:R) is the main limiting factor; integrity impact is low while availability impact is high. Severity is rated CVSS 7.1, EPSS is low at roughly 0.10% (1st percentile), no public exploit code or PoC reference has been identified at time of analysis, and the issue is not in CISA KEV; Ubiquiti has released a fixed firmware version.
Stack-based buffer overflow in WatchGuard Fireware OS lets a network-adjacent attacker execute arbitrary code with root privileges by presenting a specially crafted certificate to the logging service. Exploitation is condition-gated: the firewall must be configured to forward logs to a remote log server, and the attacker must be able to impersonate that server or intercept traffic to it. No public exploit was identified at the time of analysis, and the vendor has released fixed versions.
A stack buffer overflow in Asterisk's res_xmpp module allows an XMPP-authenticated attacker who can reach the PBX's configured XMPP connection to crash the Asterisk process, giving a high availability impact (CVSS 3.1 AV:N/AC:L/PR:L/UI:N/A:H, 7.1). Only deployments where res_xmpp is loaded and an XMPP connection is defined in xmpp.conf are exposed - neither is part of a default configuration - and the attacker must hold a valid account on the same or a federated XMPP server. Fixed releases are available for all supported branches (20.20.1, 21.12.3, 22.10.1, 23.4.1, plus Certified 20.7-cert11 and 22.8-cert3); no public exploit has been identified at time of analysis and the issue is not in CISA KEV.
Improper output sanitization in Tanium Enforce 2.9, 2.10, and 3.0 allows a network-reachable attacker to craft malicious output that can lead to limited information disclosure and integrity impact when a legitimate user views or exports affected data (CVSS 7.1). The CVSS vector indicates no required privileges (PR:N) but user interaction (UI:R) is required, and whether injecting the payload itself requires prior authenticated access to Enforce is unverified; no public exploit identified at time of analysis. Tanium has released fixed builds in Enforce 2.9.736, 2.10.778, and 3.0.373, and the vuln.today priority engine rates remediation as 'this_week'.
Improper output sanitization in Tanium Comply allows a remote, unauthenticated attacker to introduce crafted content that is rendered or exported when an authorized operator views the affected output, with the scope change indicating impact beyond the attacker-controlled component, such as the operator's session or an exported file. Only Comply 2.32 before 2.32.263, 2.35 before 2.35.323, and 2.37 before 2.37.326 are affected; confidentiality, integrity, and availability impacts are Low, and the required user interaction is the main limiting factor. No public exploit was identified at the time of analysis, and patched builds are available from Tanium.
Server-side request forgery (SSRF) in NVIDIA TensorRT-LLM 1.3.0rc24 allows an attacker who can reach a local input vector to influence server-side requests and potentially disclose information. NVIDIA rates the issue CVSS 7.1 with a local attack vector (AV:L), no authentication required once that vector is reached, no user interaction, changed scope, and high confidentiality impact only. No public exploit code or active exploitation is identified in the available intelligence, and applicability to versions other than 1.3.0rc24 is not confirmed.