Remote Code Execution

Out-of-bounds write in Adobe Lightroom Classic allows arbitrary code execution under the context of the logged-in user when a victim opens a specially crafted file. All releases from 15.2 through 15.4.1 are confirmed affected; Adobe has released version 15.5 as the fix per advisory APSB26-94. No public exploit code or CISA KEV listing has been identified at time of analysis, but the combination of a concrete patch release and high CVSS impact warrants prompt remediation for any organization with creative or photography workflows.

Out-of-bounds write in Adobe Lightroom Classic versions up to 15.4.1 enables arbitrary code execution in the security context of the current user through a maliciously crafted file. The flaw, classified under CWE-787 (Memory Corruption / Buffer Overflow), resides in file-parsing logic and requires no elevated privileges on the attacker's side - only that a victim opens the weaponized file. No public exploit code or active exploitation has been confirmed at time of analysis; vendor-released patch 15.5 is available.

Out-of-bounds write in Adobe Lightroom Classic versions up to 15.4.1 enables arbitrary code execution in the context of the logged-in user when a victim opens a specially crafted file. Tagged as Memory Corruption, RCE, and Buffer Overflow by Adobe PSIRT, the vulnerability carries a CVSS 7.8 (High) rating. No public exploit or CISA KEV listing has been identified at time of analysis, but the file-based attack vector makes it suitable for phishing and malicious download campaigns targeting creative professionals.

Out-of-bounds write in Adobe Lightroom Classic versions up to 15.4.1 enables arbitrary code execution in the context of the logged-in user when a crafted malicious file is opened. The vulnerability is rooted in improper memory boundary validation during file parsing (CWE-787), yielding full confidentiality, integrity, and availability impact per CVSS. No public exploit or CISA KEV listing has been identified at time of analysis, but the RCE-class severity and file-delivery vector make this a meaningful phishing or watering-hole risk for creative professionals.

Out-of-bounds write in Adobe Lightroom Classic versions up to and including 15.4.1 allows arbitrary code execution in the context of the logged-in user when a specially crafted file is opened. Adobe's PSIRT reported the issue under advisory APSB26-94, and the ENISA EUVD tracks it as EUVD-2026-56829. No public exploit or CISA KEV listing has been identified at time of analysis, though the CWE-787 memory corruption class combined with a confirmed high C/I/A impact triad warrants prompt patching across all affected release branches.

Unsafe deserialization of untrusted data in Adobe Lightroom Classic (versions ≤15.4.1) enables arbitrary code execution in the context of the authenticated user when a victim opens a specially crafted file. Classified as CWE-502, this file-format exploit carries a CVSS 3.1 score of 8.6 with changed scope (S:C), indicating impact extends beyond the Lightroom process itself to broader user-accessible system resources. No public exploit identified at time of analysis and no CISA KEV listing; Adobe has released a fix in version 15.5 via security bulletin APSB26-94.

SQL injection in Adobe Campaign Classic (ACC) v7 through 7.4.3 build 9399 allows a remote attacker to inject crafted SQL and achieve arbitrary code execution in the context of the current user, per Adobe advisory APSB26-123. The flaw is rated CVSS 9.0 (critical) with a scope change, but Adobe notes exploitation depends on conditions beyond the attacker's control (AC:H). No public exploit identified at time of analysis and it is not listed in CISA KEV.

Arbitrary code execution in Adobe Lightroom Classic versions up to and including 15.4.1 is achievable by tricking a victim into opening a specially crafted malicious file, exploiting an integer overflow in the application's file parsing logic (CWE-190). The flaw yields full compromise of confidentiality, integrity, and availability within the current user's context, consistent with the CVSS 7.8 score. Adobe has released a patch in version 15.5 per advisory APSB26-94; no public exploit has been identified at time of analysis.

Arbitrary code execution in Adobe Campaign Classic (ACC) v7 through 7.4.3 build 9399 lets a remote, unauthenticated attacker run code in the context of the current user via an incorrect-authorization flaw (CWE-863). The maximum CVSS 3.1 base score of 10.0 reflects a network-reachable, no-interaction bug with a changed scope, meaning impact can extend beyond the initially vulnerable component. There is no public exploit identified at time of analysis and it is not listed in CISA KEV, but the combination of unauthenticated RCE and a vendor PSIRT advisory makes this a top-priority patch.

Code injection (CWE-94) in Microsoft PowerShell enables a local, unauthenticated attacker to bypass a security feature through user interaction, resulting in high-impact compromise of confidentiality, integrity, and availability. The CVSS 3.1 score of 7.8 reflects a local attack vector requiring user interaction, with no privileges needed from the attacker. No public exploit code or CISA KEV listing has been identified at time of analysis, but the RCE tag and full C/I/A impact make this a priority for environments running PowerShell on Windows.

Remote code execution in Microsoft Visual Studio Code allows a network-based attacker to inject and execute arbitrary code on a victim's machine with no prior authentication, provided the victim performs a user interaction - such as opening a malicious workspace, file, or URI. Classified under CWE-94 (Code Injection), the vulnerability carries a CVSS base score of 8.8, reflecting high impact across confidentiality, integrity, and availability. No public exploit code or CISA KEV listing was identified at the time of analysis, but the attack complexity is rated Low, making reliable exploitation straightforward once a victim is lured.

Code injection (CWE-94) in Microsoft SharePoint Server exposes authenticated network users to high-impact compromise across confidentiality, integrity, and availability on unpatched servers running SharePoint 2016 Enterprise, 2019, or Subscription Edition. The CVSS 8.8 rating (AV:N/AC:L/PR:L) reflects network exploitation requiring only low-privileged credentials, though the CVE description characterizes the primary effect as spoofing rather than full remote code execution - a tension that deserves verification against the Microsoft MSRC advisory. No public exploit code and no CISA KEV listing have been identified at time of analysis, and SSVC rates exploitation as none and automatable as no, moderating the immediate response urgency.

Remote code execution in Windows Routing and Remote Access Service (RRAS) allows unauthenticated network attackers to fully compromise affected hosts across an exceptionally broad range of Windows versions, from Server 2012 through Server 2025 and Windows 10/11. The root cause is a use-after-free memory corruption flaw (CWE-416) combined with an authentication bypass, enabling an attacker to execute arbitrary code without credentials. No public exploit code or CISA KEV listing has been identified at time of analysis, but the network-exposed attack surface, zero-authentication barrier, and breadth of affected platforms make this a high-priority patching target. A vendor-released patch is available via Microsoft MSRC.

Incorrect Authorization (CWE-863) in Adobe ColdFusion 2023 and 2025 allows adjacent-network attackers to bypass authorization controls and execute arbitrary code in the context of the ColdFusion service account. The vulnerable administrative component is confined to an administrative network zone by default, limiting internet-scale exposure but leaving internal networks and misconfigured deployments at significant risk. No public exploit code or CISA KEV listing has been identified at time of analysis, though the complete CIA impact and zero-authentication requirement make this a high-priority patch target for any organization running ColdFusion.

Heap-based buffer overflow in Adobe ColdFusion 2023 and 2025 enables unauthenticated remote attackers to execute arbitrary code in the context of the ColdFusion server process. The vulnerability requires high-complexity exploitation conditions outside the attacker's direct control, yielding a CVSS score of 8.1 rather than critical, but the triple-high impact (full confidentiality, integrity, and availability compromise) makes it a high-priority target. No public exploit code or CISA KEV listing has been identified at time of analysis, though ColdFusion's historical targeting by ransomware and state-sponsored actors elevates real-world urgency beyond what the AC:H qualifier alone would suggest.

OS command injection in Adobe ColdFusion 2023 and 2025 enables remote, unauthenticated attackers to execute arbitrary operating-system commands in the context of the ColdFusion service account. Rated CVSS 10.0 with a scope change (S:C), this flaw stems from improper neutralization of special elements passed into an OS command and requires no user interaction. No public exploit identified at time of analysis, but ColdFusion's history as an attacker-favored target makes rapid patching essential.

Cross-site scripting in Adobe ColdFusion 2023 and 2025 enables arbitrary code execution in the context of an administrative user when a victim opens a crafted malicious file. The vulnerable component is restricted to the administrative network zone by default (AV:A), limiting exposure to internal or VPN-connected segments, but the scope change (S:C) with full CIA impact confirms that successful exploitation can cascade beyond the immediate ColdFusion application context. No public exploit has been identified at time of analysis; however, Adobe's advisory APSB26-90 confirms the RCE potential and the high severity rating of 8.8.

Privilege escalation in Intel PROSet/Wireless WiFi Software for Windows exposes local systems to code execution by unprivileged attackers via a Ring 2 (device driver layer) improper access control flaw. Any local process with no privileges required and no user interaction can leverage the driver's failure to enforce access boundaries to achieve local code execution, with follow-on high confidentiality impact on the broader system - consistent with a kernel-level scope change. No public exploit identified at time of analysis, and no CISA KEV listing; however, the low complexity and zero-privilege requirement make this a credible post-compromise escalation vector for threat actors with any form of local code execution already established.

Remote code execution in Zoom Clients' annotator function allows a malicious meeting participant to compromise another participant's device by exploiting a use-after-free memory corruption flaw during a live meeting session. The CVSS:3.1/AV:N/AC:H/PR:N/UI:R/S:C/C:H/I:H/A:H vector confirms network-accessible cross-participant exploitation with full CIA impact and scope change, requiring no in-meeting privileges but demanding high attack complexity consistent with UAF heap manipulation. No public exploit code has been identified at time of analysis, and no CISA KEV listing was found.

OS command injection in Google Turbinia's forensic worker fleet allows an attacker who can submit processing requests or control evidence path/name values to execute arbitrary commands on the worker nodes. The root cause, confirmed by PR #1631, is multiple workers (grep, photorec, strings) constructing shell command strings via Python format interpolation and invoking them with shell=True, allowing shell metacharacters in evidence paths to break out of the intended command. The fix, released 2026-07-10, replaces all affected workers with list-based subprocess invocation, eliminating shell interpretation.

Remote code execution in Zoom Clients allows a malicious meeting participant to exploit a buffer over-write vulnerability in the annotator function, achieving code execution on another participant's device over the network. The CVSS 3.1 vector (AV:N/AC:H/PR:N/UI:R/S:C/C:H/I:H/A:H) reflects high-complexity exploitation with a changed scope, meaning impact extends beyond the Zoom process to the underlying host. No public exploit code and no CISA KEV listing have been identified at time of analysis, but the RCE potential and network vector make this a high-priority patch target for organizations relying on Zoom for sensitive communications.

Code injection in Vim's vimball plugin prior to version 9.2.0847 allows arbitrary operating-system command execution when a user installs a crafted vimball archive and later uninstalls it. An attacker who delivers a malicious vimball containing a member named .VimballRecord can overwrite the installation record with arbitrary Ex commands - including shell invocations via :! - which execute with the victim user's full privileges when vimball#RmVimball() is subsequently called. No public exploit has been identified at time of analysis, and no active exploitation is confirmed in CISA KEV; the upstream patch is available as Vim 9.2.0847.

Code injection in the Perl DBI module as packaged for Red Hat Enterprise Linux allows a network-accessible, low-privileged attacker to execute arbitrary code, resulting in full confidentiality, integrity, and availability compromise. This CVE is a remediation of an incomplete fix for CVE-2026-14380, specifically targeting RHEL 9.8.z and 10.2.z patch levels where the prior partial fix left an exploitable residual path. No public exploit or CISA KEV listing is identified at time of analysis, but the CWE-94 classification and RCE tag indicate high potential impact if exploitation conditions are met.

Unauthenticated remote code execution in Dassault Systèmes SIMULIA Execution Engine (Release 2023 through Release 2026) allows a network attacker to run arbitrary code on the server by submitting a malicious serialized object that the application deserializes without validation. The flaw carries a maximum CVSS 3.1 base score of 10.0 and requires no privileges or user interaction, making any exposed instance fully compromisable. No public exploit code has been identified at time of analysis and it is not listed in CISA KEV, but the CWE-502 root cause combined with a network attack vector makes it a high-priority patching target.

Remote code execution in Streambert (Electron desktop streaming/download app) versions before 2.5.0 lets a compromised renderer process coerce the trusted main process into fetching and executing an arbitrary binary through an unvalidated auto-updater URL. Any attacker who first gains control of the renderer - for example via malicious streamed web content or an XSS-style flaw - can escalate that foothold into full code execution on the host. Fixed in 2.5.0; no public exploit identified at time of analysis and the issue is not listed in CISA KEV.

Unauthenticated remote code execution affects Siemens SIMATIC IoT2050 Advanced (6ES7647-0BA00-1YA2) gateways running Industrial OS with Node-RED installed, in all versions prior to V4.3.4.1. The Node-RED HTTP interface ships without authentication, exposing programming nodes that run system commands, so any attacker with network reachability can push malicious flows and execute code as the highest-privilege user. There is no public exploit identified at time of analysis and it is not in CISA KEV, but the CVSS 10.0 rating and trivial exploitation make this a top-priority patch for exposed devices.

Remote code execution in Craft CMS allows an authenticated control panel user to escape the Twig sandbox and execute arbitrary server-side PHP, even when sandbox mode is explicitly enabled via `enableTwigSandbox()`. The root cause is that the sandbox's class allowlisting propagates through the entire class hierarchy of `ElementInterface`, ultimately exposing the well-known `yii\base\Component` arbitrary function-call gadget - the identical primitive behind at least four prior Craft CMS CVEs. Affected versions are 5.0.0-RC1 through 5.10.6 and 4.0.0-RC1 through 4.18.2; patched releases 5.10.7 and 4.18.3 are available, and no CISA KEV listing exists at time of analysis.

Authenticated remote code execution in Craft CMS control panels allows any user with a valid panel session to execute operating system commands as the PHP/web user by exploiting a config-cleanse bypass in element-search condition handling. Affected versions span the 4.x branch from 4.0.0-RC1 through 4.18.1, and the 5.x branch from 5.0.0-RC1 through 5.10.5. The GHSA advisory confirms lab-verified exploitation; no public exploit code has been identified at time of analysis, and this CVE does not appear in the CISA KEV catalog.

Remote code execution in n8n's Git node allows authenticated users with workflow creation and execution rights to run arbitrary OS commands as the n8n process user by staging a crafted repository that triggers git hook execution under default git security settings. All 1.x versions before 1.123.67, 2.x before 2.31.5, and 2.32.x before 2.32.1 are affected across both self-hosted and cloud deployments. No public exploit or CISA KEV listing has been identified at time of analysis, but the low-complexity attack path and broad version coverage make this a high-priority patch for any deployment permitting untrusted authenticated users.

Expression sandbox escape in n8n's workflow engine allows any authenticated user with workflow creation or modification permissions to execute arbitrary operating system commands on the host running n8n. The flaw, classified as CWE-94 code injection, affects all npm releases of n8n-io/n8n prior to version 2.31.5 and all 2.32.x releases prior to 2.32.1. No public exploit has been identified at time of analysis, but the CVSS 4.0 score of 8.7 and the full confidentiality, integrity, and availability impact ratings reflect a genuine high-severity remote code execution primitive for any n8n deployment where workflow editing rights are not tightly controlled.

Root-level OS command injection in wg-easy 15.3.0 lets a low-privileged web user holding the clients.create permission escalate to full host compromise. Because the client name is written verbatim into the WireGuard config, an attacker embeds newline-separated PostUp directives that wg-quick later executes as root. No public exploit identified at time of analysis, but the CVSS 9.9 (S:C, full C/I/A) and simple, GUI-driven injection make this a high-priority issue for any exposed wg-easy instance.

Remote code execution in ZoneMinder 1.39.17 and earlier allows any authenticated user - regardless of assigned role - to execute arbitrary OS commands by exploiting a broken permission check in the Filter class. The flaw arises because `canEdit()` and `canDelete()` invoke nonexistent methods on the `ZM\User` class, causing PHP's `__call()` magic method to silently return a truthy value and bypass all role-based authorization for filter operations. No public exploit code or CISA KEV listing has been identified at time of analysis, but the extremely low exploitation bar - any valid account suffices - makes this a high-priority remediation target for all ZoneMinder deployments.

Remote code execution in Fusio 8.8.3 allows authenticated Developer-role users to escape the PHP-Sandbox and execute arbitrary OS commands on the server. The root cause is a CWE-78 OS Command Injection flaw: the sandbox allow-list permits PHP functions that transitively invoke system(), enabling full sandbox escape. With a valid Developer-role account, an attacker can achieve complete server compromise. No public exploit code and no CISA KEV listing have been identified at time of analysis, though the low attack complexity and network accessibility elevate priority for any Fusio deployment where Developer accounts are granted to untrusted parties.

Remote code execution in PrefectHQ Prefect through version 3.8.2 allows authenticated users to inject arbitrary git arguments via the unsanitized `branch` field of the `git_clone` pull step, enabling arbitrary command execution on the Prefect server. This vulnerability is a distinct bypass of the incomplete fix applied for CVE-2026-5366, meaning the earlier remediation left the underlying code path exploitable by a different injection technique. No public exploit or CISA KEV listing has been identified at time of analysis, but the low-complexity network vector and authenticated-user scope make this high-priority for any organization running self-hosted Prefect orchestration infrastructure.

Unauthenticated remote code execution in SAP Commerce Cloud's Data Hub Adapter lets attackers abuse a default authentication client to send specially crafted input to functions that lack sufficient validation, yielding arbitrary code execution and a scope change that compromises internal components. This is a maximum-severity CVSS 10.0 flaw (CWE-94 code injection) affecting confidentiality, integrity, and availability. No public exploit is identified at time of analysis, but the trivial attack profile (network, low complexity, no authentication, no user interaction) makes it a top-priority patch.

Remote code execution in Red Hat JBoss Enterprise Application Platform (EAP) via the openjdk-orb CORBA bridge allows unauthenticated network attackers to load and instantiate arbitrary Java classes from attacker-controlled URLs during IIOP object unmarshalling. The attack targets port 3528 and executes before EJB security interceptors are invoked, meaning EAP's own authorization layer provides no protection once the malicious payload is in flight. No public exploit or CISA KEV listing is confirmed at time of analysis, but the CWE-829 root cause class and full CIA:HHH CVSS impact reflect genuine arbitrary-code-execution potential under the required configuration.

Remote code execution in Crafty Controller (crafty-4), a web-based management panel for Minecraft servers, arises from path traversal in the server-import and admin file-upload features that lets an authenticated user write files to arbitrary filesystem paths accessible to the application. By planting a file in an executable or auto-loaded location, a low-privileged authenticated attacker can escalate to code execution in the context of the Crafty Controller process. No public exploit has been identified at the time of analysis, and the issue is not listed in CISA KEV; EPSS data was not provided.

Remote code execution via improper input validation in a VAPIX API parameter on Axis OS affects Axis network camera and edge device deployments. An attacker holding administrator-privileged service account credentials can send a crafted API parameter to trigger the flaw, execute arbitrary code on the device, and potentially escalate to higher privileges. No public exploit code has been identified at time of analysis, and this vulnerability is not listed in the CISA KEV catalog; however, the full-impact CVSS score of 7.2 with network attack vector and high C/I/A ratings reflects the severity of a successful post-authentication compromise.

OS command execution in SAP Manufacturing Integration and Intelligence (MII) lets an already-highly-privileged application user submit crafted input to affected functionality that is passed to the operating system without sufficient validation (CWE-94 code injection). Because exploitation crosses from the application into the underlying host (scope change), a successful attack yields full confidentiality, integrity, and availability impact on the server, not just the app. The issue was reported by SAP and carries a 9.1 CVSS; there is no public exploit identified at time of analysis and it is not listed in CISA KEV.

Server-side template injection in ERPNext's Process Statement of Accounts feature (versions prior to 15.118.0 and 16.29.0) lets an authenticated user with a common operational role inject Jinja template expressions into the subject, body, and pdf_name fields, which are rendered with unrestricted globals including frappe.utils, yielding arbitrary server-side code execution. Tagged RCE/SSTI (CWE-1336) and rated CVSS 9.9 with scope change; no public exploit identified at time of analysis, but the trivial exploitation path and low privilege requirement make this a high-priority patch.

Privilege escalation in Red Hat OpenShift AI's training-operator enables any user holding a standard Kubernetes edit or admin role in any cluster namespace to break out of namespace isolation, impersonate service accounts, access the host filesystem, and achieve remote code execution. The root cause is a design flaw: the operator aggregates training job permissions onto native Kubernetes built-in ClusterRoles, and allows unrestricted PodTemplateSpec passthrough - meaning users can submit training jobs containing arbitrary pod configurations such as hostPath mounts or elevated service account tokens. No public exploit code or CISA KEV listing was identified at time of analysis, but the low access complexity and the broad availability of edit/admin roles in typical ML platform deployments make this an urgent remediation priority.

Privilege escalation in Red Hat OpenShift AI (RHOAI) training operator overlay allows any namespace editor to manage TrainJob resources beyond their intended authorization scope. The RHOAI overlay incorrectly injects trainjobs CRUD permissions into the native Kubernetes edit ClusterRole, violating the principle of least privilege. When chained with a companion flaw (TRN-01) that permits arbitrary pod configuration within TrainJobs, a remote attacker holding namespace editor credentials could escalate to arbitrary code execution - no public exploit is identified at time of analysis, and KEV listing has not been confirmed.

Unauthenticated remote code execution affects Feast (the open-source ML feature store), where user-defined functions stored in the registry are serialized with Python's 'dill' library and deserialized without validation. In default configurations a remote attacker can store a malicious UDF and achieve unauthenticated code execution on the feature server; an authenticated attacker can bypass authorization during deserialization to run code on the registry server, enabling cross-tenant data access and lateral movement. Rated CVSS 9.9 with a scope change; no public exploit has been identified at time of analysis, and it is not listed in CISA KEV. Feast is bundled in Red Hat OpenShift AI (RHOAI), which is the reporting vendor's affected product.

Feast and feast-operator ship with a default 'no_auth' security configuration, leaving the feature-server, registry-server, and offline-server endpoints completely unauthenticated and accessible to any network-reachable attacker in Red Hat OpenShift AI (RHOAI) deployments. Exploitation enables three distinct high-impact outcomes: remote code execution via injection of a malicious User-Defined Function (UDF) onto the feature-server, denial of service by forcing full re-materialization of all tenant features, and unauthorized read access to cross-tenant feature data. No public exploit code has been identified at time of analysis and CISA KEV status is not confirmed, but the trivial exploitation conditions - no authentication, no interaction, default configuration - make this a realistic priority for any RHOAI operator.

Privilege escalation in Red Hat Data Science Pipelines (DSP) allows an authenticated namespace editor to gain node-root access by submitting a crafted Argo Workflow via the V1 API path. The API server, acting as a confused deputy (CWE-266), creates pods with elevated privileges on the attacker's behalf, effectively breaking out of Kubernetes namespace isolation. Successful exploitation enables arbitrary code execution and full control over the underlying cluster node; no public exploit has been identified at time of analysis, and this is not listed in CISA KEV.

Arbitrary code execution in Red Hat OpenShift AI's trustyai-service-operator enables authenticated cluster users to bypass pod security policies via crafted sidecar container configurations in LMEvalJob resources. The CVSS 8.1 (High) score reflects network-accessible exploitation requiring only low-privilege cluster authentication, with high confidentiality and integrity impact against the affected workload context. No public exploit code has been identified and no CISA KEV listing exists at time of analysis, but the low barrier of entry (any authenticated cluster user) elevates insider-threat and compromised-account risk for RHOAI deployments.

Remote code execution in Cachet through version 2.4.1 allows authenticated low-privilege users to inject Blade directives or Twig-style filters into incident templates, which are interpreted and executed as arbitrary PHP by the server when incidents are rendered. Any account with permission to create or edit incident templates can run OS commands as the web server process, yielding full compromise of the hosting environment. A publicly available proof-of-concept has been confirmed by VulnCheck and the upstream GitHub issue; no active exploitation has been recorded in CISA KEV at time of analysis.

Authenticated remote code execution in CyberPanel 2.4.3 allows any logged-in user to escalate to persistent root-level SSH access by exploiting command injection in the remote backup destination feature. The backup module concatenates user-controlled parameters-IP address, SSH port, username, and password-directly into OS-level shell commands without sanitization, and the SSH key retrieval process fails to verify host authenticity, enabling a malicious server to supply a public key that CyberPanel writes to /root/.ssh/authorized_keys. A publicly available proof-of-concept demonstrates the complete exploit chain; the vendor has released a fix in upstream commit eca0c3c, though no tagged release version has been independently confirmed.

Remote OS command execution in SPIP before 4.4.18 is achievable by any authenticated editor-level user on SQLite-backed installations via the navigation menu endpoint. Array-typed GET parameters submitted to /ecrire/?exec=navigation bypass the CMS sanitization layer and break out of an internal PHP string evaluation context, enabling arbitrary command execution in the web server process. MySQL-backed deployments are not affected; no public exploit has been identified at time of analysis.

Arbitrary command execution in the goose AI agent's `goose review` command allows a malicious Git repository to run attacker-controlled code on the host before any model interaction occurs. The vulnerability exists in versions prior to 1.44.0 and stems from goose invoking the system `git` binary without suppressing the `core.fsmonitor` configuration key, meaning a repository's `.git/config` can nominate an arbitrary command that Git then executes during index refresh. Exploitation requires only that a user runs `goose review` inside a malicious repository, and the injected command inherits the user's full environment - including provider API keys and filesystem access. No public exploit has been identified at time of analysis, though the fix commit and advisory are public.

Arbitrary code execution in GIMP's PSD import plugin allows an attacker to run code in the context of the user who opens a maliciously crafted .psd file. An unsigned integer underflow in the plugin's block_rem length variable (CWE-191) corrupts parsing of Photoshop layer resource blocks, letting attacker-controlled data be treated as trusted structures. Reported by Red Hat and tracked upstream in GNOME's GIMP tracker; SSVC lists a proof-of-concept, though it is not in CISA KEV and EPSS exploitation probability is low (0.31%).

Heap overflow in GIMP's Seattle Filmworks file loader allows an attacker who tricks a user into opening a specially crafted SFW file to write several kilobytes of attacker-controlled data beyond the intended buffer, resulting in memory corruption that can lead to arbitrary code execution or denial of service. Affected platforms confirmed via CPE data are Red Hat Enterprise Linux 6 through 9 with GIMP installed. No public exploit code has been identified and this CVE is not listed in CISA's KEV catalog, but the high CVSS 7.8 score reflects the full confidentiality, integrity, and availability impact achievable once a victim is socially engineered into opening the malicious file.

Remote code execution in NASA fprime-gds through version 3.4.3 lets an unauthenticated remote attacker fully compromise the ground data system and inject arbitrary commands to any connected spacecraft. The Flask ground station exposes every endpoint without authentication (CWE-306), while a path traversal in the upload/download handler permits reading and writing arbitrary files, and a hardcoded Flask secret key (CWE-321) enables session forgery; chained, they yield code execution on the ground station host. No public exploit is identified at time of analysis, but the CVSS 9.8 rating and unauthenticated network vector make this a high-priority, real-world risk for anyone running F Prime ground software.

Remote code execution in Apache Ranger versions up to and including 2.8.0 arises from JDBC URL injection (CWE-94), letting an attacker who can influence a database connection string coerce the server into loading attacker-controlled JDBC driver behavior and executing arbitrary code on the Ranger host. This flaw affects the centralized authorization framework used to govern access across Hadoop-ecosystem services, so compromise undermines the very control plane meant to enforce data security. There is no public exploit identified at time of analysis and EPSS is low (0.36%, 29th percentile), but the fix in 2.9.0 and the critical 9.8 CVSS warrant prompt patching.

Remote code execution in Apache Ranger 2.8.0 and earlier arises from arbitrary Java class instantiation in the plugin-schema-registry component, letting an attacker who can reach the affected interface load and run attacker-controlled classes and achieve full compromise (CWE-94). The flaw carries a critical CVSS 9.8 and is fixed in Ranger 2.9.0. There is no public exploit identified at time of analysis and it is not listed in CISA KEV, with EPSS at a modest 0.58% (45th percentile).

Remote code execution in Apache Ranger versions up to and including 2.8.0 stems from unsafe script evaluation in the GraalScriptEngineCreator component, letting an attacker inject and execute arbitrary code (CWE-94) on the Ranger host. Because Ranger is the central authorization and data-governance layer for Hadoop-ecosystem services, compromise can undermine access control across an entire big-data platform. There is no public exploit identified at time of analysis, it is not on CISA KEV, and EPSS is modest (0.34%, 26th percentile), but the 9.8 CVSS rating reflects the potential for full host compromise once conditions are met.

Remote code execution in the Fabrik application-builder extension for Joomla (versions before 4.6.7) allows unauthenticated attackers to execute arbitrary code by abusing the ajax_calc endpoint of the bundled calc form plugin. The calc plugin evaluates user-influenced expressions server-side, and the AJAX handler exposes that evaluation without authentication, giving remote attackers full control of the underlying PHP process. CVSS 4.0 rates this 10.0 (maximum); no public exploit has been identified at time of analysis, and the CVE is not listed in CISA KEV.

Stack-based buffer overflow in libril_sem.so - Samsung's Radio Interface Layer shared library - enables privileged local attackers to execute arbitrary code on affected Samsung Mobile Devices prior to SMR Aug-2026 Release 1. The CVSS 4.0 vector (AV:L/PR:H) confirms exploitation requires established local access with high privileges, limiting the realistic attacker pool to already-compromised or insider-threat scenarios. No public exploit code or CISA KEV listing has been identified at time of analysis.

Unauthenticated remote code execution in the Product Input Fields for WooCommerce WordPress plugin before 2.0.2 arises because the plugin skips file-type validation whenever its accepted-types setting is left blank - a configuration the plugin's own documentation advertises as 'accept all files.' Attackers can upload arbitrary files, including PHP webshells, and execute them on servers whose upload directory does not enforce access restrictions. Publicly available exploit code exists (WPScan), though the EPSS score is low (0.20%, 10th percentile) and it is not listed in CISA KEV.

Remote code execution in the Squeeze WordPress plugin (versions prior to 1.7.12) is achievable by any authenticated user holding the upload_files capability - granted by default to the Author role and above. The plugin's attachment-update action writes per-size image data to the uploads directory without validating file type or extension, permitting a PHP webshell to be placed in a web-accessible path and directly invoked via HTTP. A publicly available proof-of-concept exists per WPScan; no public exploit identified at time of analysis in CISA KEV, and EPSS is low at 0.19% (9th percentile), suggesting limited observed in-the-wild exploitation despite the public PoC.

Heap out-of-bounds write in GStreamer's adpcmdec element allows a crafted WAV file containing malformed multi-channel IMA/DVI ADPCM audio to corrupt application memory, cause denial of service, or potentially achieve arbitrary code execution. Affected deployments include Red Hat Enterprise Linux 7 through 10, where the vulnerable gst-plugins-bad package is distributed. No confirmed active exploitation (not in CISA KEV) and no public proof-of-concept has been identified at time of analysis, though the CVSS score of 7.6 reflects the realistic network-deliverable, user-interaction-dependent attack path.

Full network takeover via authentication bypass in the InfiniteWP Client WordPress plugin before 1.13.6 affects WordPress Multisite installations, where the plugin fails to validate the site-connection state and the authenticity of requests to its remote-management endpoint. Unauthenticated remote attackers can bind their own management key, hijack an administrator session, and gain control of every site in the network, culminating in remote code execution. No public exploit identified at time of analysis; EPSS is low at 0.19% (9th percentile) despite the maximal 9.8 CVSS, and a vendor patch is available.

Command injection via unsanitized backup filenames in Kakoune (all versions prior to 2026.05.21) enables arbitrary Kakoune script and shell command execution when a victim opens a file in the editor. The bundled `autorestore.kak` script - enabled by default - discovers backup files by pattern-matching filenames in the same directory as the opened file, then incorporates those filenames directly into shell command strings without validating special characters in the suffix component. No public standalone exploit has been identified at time of analysis, though the fix commit (25c7b13) includes test cases that explicitly demonstrate the injection pattern using filenames such as `.out.kak.;"foo"`, confirming exploitability.

Cross-origin request forgery leading to remote code execution in OpenYak before 1.1.3 lets any malicious webpage a user visits drive the desktop backend's loopback HTTP API (typically 127.0.0.1:19141) to run arbitrary shell commands on the host, terminate the service, and steal chat history and account PII. The local API ships with no Origin validation, no loopback authentication, no Content-Type enforcement, and a wildcard CORS policy, so the victim's browser is abused as a proxy that defeats OS-level network isolation. No public exploit identified at time of analysis; the flaw is fixed in version 1.1.3.

Host root code execution in Kata Containers kata-runtime prior to 4.0.0 lets a pod user escape the guest VM boundary by abusing the unvalidated 'io.katacontainers.config_path' pod annotation, which caused the runtime to load an arbitrary host TOML configuration file without restriction. Because that config selects the hypervisor and virtio-fs daemon binaries, an attacker who can both set the annotation and drop a file at a host-visible path can point the runtime at an attacker-controlled binary and have it executed as root on the host, collapsing the security isolation Kata exists to provide (CVSS 9.6, scope-changed). There is no public exploit identified at time of analysis, but the fix (removal of the annotation in version 4.0.0) is confirmed by the vendor GHSA advisory and commit.

Remote code execution in ZenML 0.94.6 allows any attacker with write access to a shared artifact store to plant a malicious cloudpickle payload that executes arbitrary system commands when a legitimate user or pipeline materializes the tampered artifact. The root cause is unsanitized deserialization via cloudpickle.load() in CloudpickleMaterializer without any integrity verification of the stored artifact.pkl file before loading. No public exploit code has been identified at time of analysis and the vulnerability is not listed in CISA KEV, but the attack is conceptually straightforward for any attacker with artifact store write access in a multi-user ZenML deployment.

Remote code execution in Sonatype Nexus Repository 3 (and legacy Nexus Repository 2) is achievable by any account holding the nexus:settings:update permission, which can submit arbitrary strings as realm identifiers to an internal configuration API that performs no validation against the registered realm set. Because unrecognized identifiers are persisted and re-evaluated on every realm load through a legacy code path (CWE-470, unsafe reflection), this allows unintended code to execute inside the Nexus Repository server process and can cause a persistent authentication lockout that is invisible through the administrative UI. A vendor-released patch is confirmed in Nexus Repository 3.95.0; no public exploit or CISA KEV listing has been identified at time of analysis.

Remote code execution in Sonatype Nexus Repository 3 is achievable by any user holding the nx-datastores-update permission, who can exploit an insufficiently restricted HikariCP connection-pool property passthrough in the DataStore configuration API to inject arbitrary SQL via the connectionInitSql property. On the default H2 embedded database backend - common in non-production and smaller deployments - H2's ability to invoke Java methods from SQL allows this injection to escalate to OS-level command execution as the Nexus process user. No public exploit has been identified at time of analysis, but the H2-to-RCE escalation path is a well-understood Java ecosystem technique that lowers the practical skill bar significantly.

Denial of service with potential for remote code execution affects Tobit Laboratories' TeamDavid (David Webbox) groupware through Rollout 524, where an unauthenticated attacker can crash the server by sending a specially crafted JSON body to a vulnerable API endpoint - a payload as small as 8 characters beginning with a number triggers an out-of-bounds stack write. The reliably reachable impact is a server crash, but the underlying buffer overflow could escalate to full server compromise if a stack canary is disclosed via a separate vulnerability. No public exploit identified at time of analysis and the flaw is not listed in CISA KEV; the high CVSS 4.0 base score (9.5) reflects the worst-case RCE potential rather than confirmed code execution.

Denial of service (with conditional remote code execution potential) affects Tobit Laboratories AG's TeamDavid groupware through Rollout 524, where the Webbox endpoint "//serverClient_close.html" fails to bounds-check multiple form-data parameters. An attacker submitting over-long parameter values triggers an out-of-bounds write (CWE-787) that crashes the server; if a stack canary can be leaked via a separate flaw, the memory corruption could escalate to full server compromise. No public exploit identified at time of analysis, and this CVE is one of 22 reported in the David platform by researchers at InfoGuard/NCSC.ch.

Denial of service, and potentially remote code execution, in Tobit Laboratories' David (TeamDavid/Webbox) collaboration suite affects all releases through Rollout 524. Multiple file-upload handlers fail to bounds-check the supplied filename, so an unauthenticated attacker who sends an upload request with an excessively long filename triggers an out-of-bounds stack write (CWE-787) that reliably crashes the server; where the stack layout is favorable or a stack canary can be leaked via a separate flaw, the same overflow may yield code execution and full server compromise. There is no public exploit identified at time of analysis and it is not listed in CISA KEV, but the flaw was reported by Switzerland's NCSC and detailed by InfoGuard Labs as one of 22 CVEs in David.

Heap type confusion in the Apache Fory C++ deserialization engine (versions 0.14.0 up to but not including 1.5.0) lets a crafted payload bypass type-compatibility checks during polymorphic smart-pointer deserialization, so an object of an incompatible type is handled as its declared base type. Because CWE-502 untrusted-data handling here corrupts memory rather than merely mis-parsing values, an attacker who controls serialized input can trigger undefined behavior ranging from denial of service to arbitrary code execution. This is rated CVSS 9.8 and is fixed in 1.5.0; there is no public exploit identified at time of analysis and EPSS is low at 0.21%.

DLL Search Order Hijacking in LUCID Vision Labs Arena SDK 1.0.80.49 on Windows enables a local low-privileged attacker to execute arbitrary code at the privilege level of the SDK process by planting a malicious DLL in a user-controlled directory present in the PATH environment variable. The SDK's dependency resolution fallback traverses PATH when a required DLL is not found locally, making the attack reliable and requiring no special tooling. No public exploit has been identified at time of analysis, but the technique is well-documented and the CVSS scope change (S:C) signals that impact can extend beyond the attacker's own privilege tier if the SDK runs under a higher-privileged account.

PHP Object Injection in the Ajax Search Lite WordPress plugin before 4.14.5 lets unauthenticated remote attackers pass untrusted serialized data into a PHP unserialize() sink, and where a suitable POP gadget chain exists in a co-installed plugin or theme, this escalates to remote code execution. Reported by WPScan with publicly available exploit code, it carries a CVSS 9.8 rating but a low EPSS of 0.22% (13th percentile), and is not listed in CISA KEV. A vendor patch is available in version 4.14.5.

Remote code execution in Subscriptions for WooCommerce WordPress plugin (before 2.0.1) is reachable by any authenticated Shop Manager through a missing capability check on a nonce-protected AJAX action that installs and activates arbitrary plugins from user-supplied slugs. Shop Managers, who by WordPress design are not granted plugin management capabilities, can exploit this gap to install a malicious plugin and achieve full server-side code execution. A publicly available proof-of-concept exists; however, EPSS is low (0.18%, 8th percentile) and the vulnerability is absent from CISA KEV, indicating exploitation remains opportunistic rather than widespread.

Out-of-bounds write in Autodesk Revit's PDF parser allows arbitrary code execution in the context of the current user process when a maliciously crafted PDF file is opened. The CVSS vector (AV:L/UI:R) confirms exploitation requires local user interaction - a victim must be socially engineered into opening a weaponized PDF. Impact scope covers full confidentiality, integrity, and availability of the running process (C:H/I:H/A:H), consistent with classic memory-corruption RCE chains. No public exploit identified at time of analysis and no CISA KEV listing, but the vendor-confirmed write primitive and RCE potential make this a credible threat for targeted attacks against AEC professionals.

Untrusted pointer dereference in Autodesk products' BMP file parser enables arbitrary code execution in the context of the current process. An attacker who can convince a user to open a specially crafted BMP file gains full code execution capability (C:H/I:H/A:H). No public exploit code has been identified at time of analysis, and CISA KEV does not list active exploitation, but the high CVSS score and confirmed RCE impact make patching a clear priority for Autodesk-heavy environments. Vendor advisory ADSK-SA-2026-0012 is the authoritative source.

Remote code execution is achievable in llama.cpp's LLaMA-Android JNI wrapper (builds b1886 through b7445) by exploiting a race condition use-after-free between `bench_1model()` and `free_1context()`, which share the `llama_context` heap object without any synchronization primitive. An attacker who wins the race can perform a heap spray with a crafted fake vtable, overwriting the vtable pointer at offset +0x30 in `llama_batch_allocr` so that `clear()` dispatches to attacker-controlled memory, achieving code execution within the Android application process. No active exploitation is confirmed (not in CISA KEV), and no public exploit code has been identified at time of analysis, though the fully documented vtable offset and exploitation technique substantially lower the development barrier for capable researchers.

Integer overflow in the llama.cpp LLaMA-Android JNI wrapper (builds b1886 through b7445) allows heap corruption through a crafted n_seq_max parameter, enabling denial of service or arbitrary code execution within Android applications that load untrusted GGUF model files. The flaw exists in the new_1batch() function, which multiplies sizeof(llama_seq_id) by an attacker-controlled value without bounds checking, causing heap allocation to wrap and undersize the buffer. No public exploit has been identified at time of analysis and this CVE is not in CISA KEV; a patched release (b7446) is confirmed available from the upstream ggml-org repository.

Out-of-bounds heap write in FFmpeg's native GoPro CineForm HD (CFHD) decoder affects versions 4.4 through 8.x, allowing remote attackers to corrupt heap memory by supplying a crafted AVI file that is opened or probed by a vulnerable build. The root failure - missing output-width bounds enforcement in the transform-type-2 reconstruction path - causes horiz_filter_clip() to write 16-bit sample rows far beyond the allocated frame buffer, with a concrete escalation path to arbitrary code execution through overwrite of a live cleanup callback pointer. No public exploit or active exploitation (CISA KEV) has been identified at time of analysis; three upstream fix commits and PR #23898 are available at code.ffmpeg.org pending a formal stable release.

Heap buffer overflow in FFmpeg's DVB subtitle parser (libavcodec/dvbsub_parser.c) affects all releases from version 0.5 through 8.x, enabling potential remote code execution when a victim processes a crafted WTV file. A signed integer overflow causes the bounds-check guard to wrap to INT_MIN, silently bypassing the PARSE_BUF_SIZE comparison and allowing attacker-controlled data to be written out-of-bounds on the heap via memcpy(). No active exploitation has been confirmed by CISA KEV and no public POC has been identified at time of analysis, but the RCE-capable heap corruption primitive in a ubiquitously embedded media library represents material risk for any pipeline that ingests untrusted media files.

SQL injection in FineAdmin V1.0 (FineAdmin.Mvc) lets remote attackers inject arbitrary SQL through the `field` and `order` parameters of paginated list endpoints, where user input is concatenated into an ORDER BY clause that cannot be safely parameterized. NVD rates it CVSS 9.8 with full confidentiality, integrity, and availability impact, and public GitHub write-ups documenting the ORDER BY-clause injection exist, though EPSS remains low at 0.40% (33rd percentile) and it is not in CISA KEV - so no public exploit is confirmed as weaponized. The 'RCE' tag reflects theoretical escalation from SQLi to code execution (e.g. via stacked queries), but the demonstrated impact is database compromise.

Unrestricted file upload in ICS-Park Smart Park Management System v2.0 lets a remote attacker upload and execute arbitrary code through the product's file upload module, yielding full server compromise (CVSS 9.8). Public exploit code exists and the flaw is automatable with total technical impact per CISA SSVC, though the EPSS probability remains low at 0.26% (17th percentile), indicating limited observed exploitation activity so far. No CISA KEV listing means no confirmed active exploitation at time of analysis.

Command injection in sysPass's backup service allows any authenticated administrator to execute arbitrary OS commands as the web server process user by storing shell metacharacters in the backup path configuration setting and triggering a backup operation. Because sysPass is a self-hosted password manager, successful exploitation grants direct read access to the master encryption key and entire credential vault, enabling mass credential disclosure and lateral movement to every system whose passwords are stored in sysPass. A public proof-of-concept is available via the referenced gist (https://gist.github.com/W40X/6747ba1b7da7bb69b0c0e162628df279); no CISA KEV listing was identified at time of analysis.

Out-of-bounds write in the Contiki-NG MQTT client (os/net/app-layer/mqtt/mqtt.c) lets a malicious or compromised MQTT broker overflow a 65-byte topic[] buffer and corrupt the adjacent payload_chunk pointer, producing an arbitrary-pointer-write primitive that can escalate to remote code execution on embedded targets lacking memory protection. The flaw stems from parse_publish_vhdr() setting topic_len_received=1 before enforcing the 64-byte topic length limit, so an over-length topic split across TCP segments bypasses the guard on parser re-entry and reaches a memcpy() that trusts the unvalidated 16-bit topic_len. Because the stack has no TLS, the plaintext channel also exposes devices to network man-in-the-middle injection; there is no public exploit identified at time of analysis and it is not in CISA KEV.

Use-after-free in the llama-server component of llama.cpp (builds b7492 through b9060) lets remote unauthenticated attackers crash the server or potentially execute code by racing six tokenization endpoints (/tokenize, /detokenize, /infill, /apply-template, /rerank, /anthropic/count_tokens) that read ctx_server.vocab directly on HTTP worker threads instead of going through the task queue. The flaw is only reachable when the server is started with --sleep-idle-seconds, since idle-driven vocab teardown on the main thread is what frees the object mid-handler. No public exploit identified at time of analysis and it is not listed in CISA KEV; the CVSS 4.0 base score is 9.2 but real-world exploitation is constrained by a hard-to-win TOCTOU race.

Remote code execution in llama.cpp's llama-server (builds b7492 through b9060) allows unauthenticated attackers to run arbitrary code by racing the server's idle-sleep transition when the --sleep-idle-seconds feature is enabled. Concurrent worker threads dereference a freed vocab pointer that can be reclaimed with attacker-controlled data. No public exploit identified at time of analysis, and the flaw is not in CISA KEV; exploitation is gated by a non-default configuration and a tight timing window (CVSS 4.0 base 9.2, AT:P/AC:H).

Heap buffer overflow in llama.cpp's server KV cache state-restore path (builds b4882 through b9058) lets an attacker who can place a crafted state file in the configured slot_save_path corrupt heap memory and potentially achieve arbitrary code execution. The flaw lives in state_read_data(), which computes a write size from an attacker-influenced cell_count without integer-overflow checking, so a malicious slot file can write past the tensor buffer. No public exploit identified at time of analysis and it is not listed in CISA KEV; the only reference is a third-party security-patches repository, so exploitation currently appears theoretical rather than active.

Integer overflow in llama.cpp's llama_batch_init() function across builds b1283 through b9058 enables heap corruption via unchecked multiplicative arithmetic in malloc() size calculations - when computed allocation sizes wrap past INT32_MAX, the resulting under-sized heap buffer is overwritten by subsequent batch operations, creating a credible path to arbitrary code execution in the context of the host process. The vulnerability affects versions 0.4.0 through 0.17.1 and any application embedding the affected build range of the library. No public exploit code has been identified at time of analysis and the CVE is not listed in the CISA KEV catalog, but the heap corruption primitive paired with the RCE tag warrants expedited patching in production AI inference pipelines.

Arbitrary file upload in WSO2 API Manager and related Carbon-based products (API Control Plane, Universal Gateway, Traffic Manager) allows an authenticated publisher-privileged user to write attacker-controlled files to arbitrary server-accessible paths via the System REST API, which can escalate to remote code execution depending on how uploaded content is served or executed. The vendor rates it CVSS 9.1 (scope-changed) and a patch is available. There is no public exploit identified at time of analysis and it is not listed in CISA KEV.