MaxKB, an open-source enterprise AI assistant by 1Panel-dev, stores user passwords as unsalted MD5 hashes, exposing all user credentials to trivial offline cracking upon any database compromise. All versions prior to 2.9.1 are affected (CPE: cpe:2.3:a:1panel-dev:maxkb). The CVSS 4.0 vector (AV:L/VC:H) confirms that while local or database-level access is a prerequisite, once hashes are obtained, full credential recovery is practically guaranteed using rainbow tables or GPU-accelerated tools such as hashcat - no public exploit identified at time of analysis, and the vulnerability is not listed in CISA KEV.
Cross-site scripting in the Joomla! Framework Filter package exposes applications to script injection through inadequately sanitized HTML attribute values in the checkAttribute methods. Authenticated users holding high-level privileges can inject malicious content that executes in the browser of any user who subsequently views the affected page. No public exploit has been identified at time of analysis, and an EPSS score of 0.01% at the 0th percentile confirms minimal current exploitation interest across the security community.
Inadequate HTML attribute sanitization in the cleanattributes() function of the Joomla! Framework Filter package exposes a cross-site scripting (XSS) vector across versions 1.0.0-3.0.5 and 4.0.0-4.0.1. An attacker holding high-privileged (administrative) access can inject malicious script content through unfiltered HTML attributes; when a victim user passively views that content, the payload executes in their browser, yielding high confidentiality impact (VC:H) consistent with session token or credential theft. No public exploit code exists and this vulnerability is not listed in CISA KEV, placing real-world risk well below what the CVSS 6.9 base score might initially suggest.
Authenticated blind SQL injection in Joomla! CMS's com_tags component allows high-privileged attackers to exfiltrate database contents by supplying maliciously crafted ORDER BY clauses that are not properly validated. Affected versions span two distinct release trains: the 4.x/5.x line (4.0.0 through 5.4.5) and the 6.x line (6.0.0 through 6.1.0), meaning the vulnerability persists across both legacy and current major releases. No public exploit code has been identified at time of analysis, and the EPSS score of 0.00% indicates negligible automated exploitation probability, though the high-privilege requirement significantly limits the realistic attacker pool.
Pre-Account Takeover in Chatwoot's OmniAuth integration affects all releases from 2.14.0 through 4.12.x, allowing an attacker who pre-registers a victim's email address to retain persistent login access after the legitimate owner authenticates via Google OAuth. The OAuth callback controller failed to invalidate attacker-set password credentials when confirming a pre-existing unconfirmed account, leaving the attacker's session viable indefinitely. No public exploit identified at time of analysis and EPSS is 0.04% (12th percentile), consistent with SSVC's 'Exploitation: none' finding, though SSVC rates technical impact as 'total' given the attacker gains full workspace access including PII, API keys, and conversation history.
OS command injection in Totolink CA750-PoE 6.2c.510 allows a low-privileged remote attacker to execute arbitrary operating system commands by manipulating the PIN argument passed to the setWiFiWpsConfig function within /cgi-bin/cstecgi.cgi. The attack requires no user interaction and is reachable over the network, making it a credible threat to any deployment exposing the device's management interface. A public proof-of-concept exploit has been published on GitHub, and EPSS places this at the 87th percentile of exploitation likelihood despite a low raw CVSS 4.0 score of 2.1 - a signal worth noting against the mismatch.
OS command injection in Totolink CA750-PoE firmware 6.2c.510 enables remote attackers with low-level credentials to execute arbitrary operating system commands via the fwUrl and magicid parameters of the recvUpgradeNewFw function within /cgi-bin/cstecgi.cgi. The firmware upgrade Setting Handler fails to sanitize user-supplied input before passing it to the underlying OS shell, exposing the device to command execution. A public proof-of-concept exploit is available on GitHub; while not yet listed in CISA KEV, the EPSS 87th-percentile ranking signals elevated real-world exploitation interest relative to the broader vulnerability landscape.
OS command injection in Totolink CA750-PoE firmware 6.2c.510 allows authenticated remote attackers to execute arbitrary shell commands by manipulating the FileName argument passed to the setUploadUserData function within the /cgi-bin/cstecgi.cgi Setting Handler. A public proof-of-concept exploit is available on GitHub, meaningfully lowering the skill barrier for adversaries. While not listed in CISA KEV, the EPSS score of 2.95% at the 87th percentile signals real-world exploitation probability well above average, making this a higher practical risk than the CVSS 4.0 base score of 2.1 alone would suggest.
OS command injection in Totolink CA750-PoE firmware 6.2c.510 allows a network-adjacent authenticated attacker to execute arbitrary operating system commands by manipulating the FileName argument passed to the setUpgradeUboot function within the /cgi-bin/cstecgi.cgi Setting Handler. Publicly available exploit code exists, hosted on GitHub, making exploitation accessible to low-skilled attackers. No public exploit identified in CISA KEV at time of analysis, though the EPSS 87th percentile ranking signals elevated exploitation interest relative to the broader CVE population despite a low absolute probability of 2.95%.
Out-of-bounds memory access in FastNetMon Community Edition through 1.2.9 allows an authenticated BGP peer to crash the monitoring process by sending a crafted UPDATE message. The parser in src/bgp_protocol.hpp correctly detects the RFC 4271 Extended Length flag but reads only one byte of the mandatory two-byte length field, silently truncating attribute lengths above 255 bytes and causing cascading misparse of subsequent BGP data. With CVSS availability impact rated High and SSVC confirming no known active exploitation, this is a targeted denial-of-service risk primarily affecting network operators using FastNetMon for DDoS detection - no public exploit code identified at time of analysis.
Missing authorization in Zyxel GS1200v3-series managed switches allows a LAN-based, unauthenticated attacker to read system configuration data from a log file by sending a crafted HTTP request to the device's web interface. All five GS1200v3 models are affected across their current firmware versions. No public exploit has been identified at time of analysis, and both EPSS (0.03%, 11th percentile) and SSVC exploitation status ('none') confirm no observed in-the-wild exploitation, placing this in the category of a real but low-urgency information disclosure risk confined to the local network segment.
Stored XSS in the Geo Mashup WordPress plugin (versions through 1.13.18) allows authenticated low-privileged users to persistently inject malicious JavaScript into pages served by the affected site. When a higher-privileged user - such as an administrator - views the content containing the stored payload, the script executes in their browser under a changed scope, enabling session hijacking, credential theft, or unauthorized administrative actions. No public exploit has been identified at time of analysis, and an EPSS of 0.03% (10th percentile) combined with SSVC exploitation status of 'none' confirms very low current real-world exploitation activity.
Out-of-bounds read in FastNetMon Community Edition through 1.2.9 exposes network monitoring infrastructure to unauthenticated remote attack via malformed NetFlow v9 UDP packets. The vulnerability resides in the options template parser (process_netflow_v9_options_template()), where attacker-controlled length fields - option_scope_length and option_length - drive iteration loops with no bounds validation, enabling reads past the end of the UDP packet buffer. With a CVSS vector of AV:N/AC:L/PR:N/UI:N and SSVC automatable:yes, mass exploitation is technically feasible, though EPSS sits at 0.03% (9th percentile) and no public exploit or KEV listing has been identified at time of analysis.
Broken access control in e107 CMS prior to version 2.3.4 permits any low-privileged authenticated user to overwrite comments authored by other users, including administrators. The server-side updateComment() function in comment_class.php accepted a comment_id from the request and issued an UPDATE SQL query filtered only by that identifier, never verifying that the requesting user owned the targeted comment. A proof-of-concept exploit exists per SSVC data, though EPSS stands at a low 0.03% (8th percentile) and no active exploitation is confirmed in CISA KEV, indicating currently limited in-the-wild activity.
Unauthenticated write access to patient electronic health records in epa4all-client 1.2.4 and earlier exposes German Telematik Infrastruktur (ePA 3.0) deployments to unauthorized data manipulation. The REST adapter component ships with no authentication or authorization controls, allowing any adjacent-network caller to write arbitrary documents to any patient EHR accessible via the institution's SMC-B card. No public exploit code has been identified at time of analysis, but the CVSS vector (AV:A/AC:L/PR:N/UI:N) confirms exploitation requires no credentials and minimal technical complexity once network-adjacent.
CSRF protection bypass in e107 CMS prior to 2.3.5 allows unauthenticated remote attackers to perform unauthorized comment moderation actions by tricking an authenticated user into visiting a malicious page. The root flaw is in session_handler::check(), which skips CSRF token validation entirely when no token is submitted, rather than rejecting the tokenless request - effectively making CSRF protection opt-in from the attacker's perspective. A proof-of-concept exists per SSVC data; the vulnerability is not listed in CISA KEV and carries a very low EPSS score of 0.01% (3rd percentile), indicating limited observed exploitation in the wild.
Unvalidated jarURI handling in Apache Flink Kubernetes Operator exposes authenticated low-privilege users to server-side request forgery and arbitrary file read primitives against the operator pod. Any Kubernetes principal holding Custom Resource create permissions can submit a malicious FlinkSessionJob CR with a crafted jarURI - pointing to local filesystem paths, internal cloud metadata endpoints, link-local addresses, or any backing store reachable through Flink's pluggable filesystem layer - and retrieve that content via the submitted job. No public exploit has been identified at time of analysis, and EPSS sits at 0.01% (3rd percentile), but the confidentiality impact is rated High by NVD given the breadth of accessible internal resources.
NVIDIA Display Driver for Linux exposes a denial-of-service condition in the Multi-Instance GPU (MIG) partition management subsystem, rooted in insecure default initialization of memory subsystem routing resources (CWE-1188). A local authenticated user - with low privileges on a Linux system running MIG-enabled Tesla, GeForce, RTX/Quadro/NVS, or Virtual GPU Manager driver branches - can trigger a hang or data corruption during partition reconfiguration, potentially disrupting all GPU workloads sharing the affected physical GPU. No public exploit code exists and this vulnerability is not in CISA KEV; EPSS sits at 0.01% (2nd percentile), indicating no observed mass exploitation at time of analysis.
Denial of service in NVIDIA Display Driver for Windows and Linux stems from improper lock management (CWE-667), where an attacker with local low-privilege access can leak held driver locks, potentially crashing or hanging the driver stack and denying GPU availability system-wide. Scope is changed (S:C), meaning the impact extends beyond the vulnerable component to the broader kernel or hypervisor layer. No public exploit identified at time of analysis and no CISA KEV listing; EPSS at 0.01% (1st percentile) and SSVC exploitation status of 'none' consistently signal low near-term exploitation likelihood.
Out-of-bounds memory read in FastNetMon Community Edition through 1.2.9 exposes sensitive process memory and can crash the collector when processing crafted NetFlow v9 packets. The NetFlow v9 data flowset parser in src/netflow_plugin/netflow_v9_collector.cpp omits the per-iteration boundary check present in the sibling Options template branch, allowing a network-adjacent attacker who can deliver UDP NetFlow v9 packets to the collector to supply malicious template definitions that drive the parser past the end of the packet buffer. No public exploit code or confirmed active exploitation has been identified at time of analysis, but a security researcher blog post at lorikeetsecurity.com documents the code-level flaw, raising the disclosure surface.
Heap-based buffer overflow in Hitachi Energy MACH HiDraw's XML parser allows a low-privileged, locally authenticated attacker to corrupt heap memory by inducing a victim to open a specially crafted XML file, with a primary impact of high availability loss (application crash) and limited confidentiality and integrity compromise. Affected versions span MACH HiDraw 9.0 through versions prior to 9.22 per EUVD-2026-31812. No public exploit identified at time of analysis, and an EPSS score of 0.01% (3rd percentile) combined with an SSVC exploitation status of 'none' confirm minimal current real-world exploitation activity.
Improper access control in Joomla! CMS com_scheduler component permits low-privileged authenticated backend users to modify the task types of existing scheduler tasks, an operation that should be restricted to higher-privileged administrators. Affected versions span Joomla! 4.1.0 through 5.4.5 and 6.0.0 through 6.1.0. While no public exploit code exists and CISA KEV has not confirmed active exploitation, the high subsequent-system impact scores (SC:H/SI:H/SA:H) in the CVSS 4.0 vector indicate that tampering with scheduler task types can produce significant integrity and availability consequences beyond the immediately vulnerable component.
Server-side request forgery in MaxKB v2.8.0 and earlier allows authenticated low-privilege users to reach internal network services by exploiting an URL parsing inconsistency at the OSS file fetch endpoint. The discrepancy between Python's urlparse validation logic and the requests HTTP client means a crafted URL can pass security checks yet still route to internal infrastructure when executed by the server. No public exploit or KEV listing exists at time of analysis; the vendor has confirmed and patched the issue in 2.8.1.
Server-Side Request Forgery in MaxKB before 2.9.1 allows authenticated users to pivot into internal network infrastructure by supplying arbitrary URLs to the work_flow_template import endpoint. The server fetches attacker-controlled URLs without validation or internal IP filtering, enabling reconnaissance of internal services, cloud metadata endpoints, or other resources unreachable from the public internet. The CVSS 4.0 score of 6.3 is driven primarily by high subsequent-system confidentiality impact (SC:H), reflecting the lateral reach into backend infrastructure that SSRF typically enables. No public exploit code or active exploitation has been identified at time of analysis.
Missing authorization controls in the WpTravelly WordPress plugin (versions through 2.1.5) allow any authenticated low-privileged user to exploit incorrectly configured access control security levels, gaining unauthorized access to functionality or data beyond their intended permissions. The flaw is network-accessible, requires no user interaction, and carries a balanced Low-CIA-triad impact per CVSS. No public exploit has been identified at time of analysis, and the EPSS score (0.03%, 10th percentile) suggests very low real-world exploitation probability despite the network-exposed attack surface.
OS command injection in haojing8312 WorkClaw up to version 0.6.4 allows a low-privileged remote attacker to bypass the application's blacklist-based command filter and execute arbitrary operating system commands. The flaw resides in the `is_dangerous` function within the Rust/Tauri agent's bash tool (`apps/runtime/src-tauri/src/agent/tools/bash.rs`), where an incomplete blacklist fails to block crafted payloads. A publicly available proof-of-concept exploit exists via a GitHub issue report; no vendor patch has been released as the project has not responded to the disclosure.
Buffer overflow in FastNetMon Community Edition through 1.2.9 allows a local attacker with no privileges to crash the FastNetMon process, disabling DDoS detection and network monitoring capabilities. The vulnerability is specifically tied to a sprintf-based overflow in the ExaBGP integration component, as documented in the Lorikeetsecurity advisory. This is one of at least three distinct buffer overflow vulnerabilities (alongside CVE-2026-48686 and CVE-2026-48689) identified in the same product version, suggesting a broader audit surfaced a class of unsafe string-handling bugs. No public exploit identified at time of analysis, and the impact is limited to availability (denial of service) with no confidentiality or integrity exposure.
Cross-site scripting in IBM Financial Transaction Manager for SWIFT Services for Multiplatforms versions 3.2.4.0 through 3.2.4.15 permits injection of arbitrary JavaScript into the Web UI, creating a pathway to credential disclosure within an authenticated operator's session. The Changed Scope (S:C) in the CVSS vector confirms the injected script executes in the victim's browser beyond the originating application boundary, elevating real-world impact in a SWIFT financial messaging context. No public exploit has been identified and EPSS stands at 0.05% (16th percentile), reflecting low immediate exploitation probability, though a discrepancy between the CVE description characterizing the attacker as unauthenticated and the CVSS PR:L vector warrants vendor clarification before finalizing risk posture.
Improper access control in JeecgBoot versions up to 3.9.1 allows authenticated low-privileged remote attackers to bypass authorization checks by manipulating the `userIdentity` argument in the SysUser component's `user.getUsername` function at the `/sys/user/login/setting/userEdit` endpoint. A publicly available proof-of-concept exploit exists via GitHub issue #9596, increasing practical risk for any multi-user JeecgBoot deployment where adversaries hold a low-privileged account. Vendor-released patch v3.9.2 is available and explicitly remediates this access control failure alongside several other high-severity issues including RCE and SSRF, indicating a broad security hardening effort in this release cycle.
Cross-site scripting in Teable's authentication redirect flow (versions 1.0-1.9.x) allows a remote unauthenticated attacker to inject and execute arbitrary JavaScript in a victim's browser by crafting a login URL with a malicious `redirect` parameter using javascript: or data: URI schemes. The vulnerable component is LoginPage.tsx in the Next.js frontend and the social auth controller adapter in the NestJS backend, neither of which validated the redirect destination before navigating. Publicly available exploit code exists (GitHub gist), but the vulnerability is not listed in CISA KEV and EPSS probability is very low at 0.04% (11th percentile), indicating no confirmed widespread exploitation.
Reflected cross-site scripting in itsourcecode Electronic Judging System 1.0 allows unauthenticated remote attackers to inject malicious scripts via the `fname` parameter of `/admin/judges.php`, executing arbitrary JavaScript in the context of a victim's browser session. The CVSS 4.0 score of 2.1 reflects the low integrity impact and mandatory user interaction, consistent with a reflected XSS that requires a victim to follow a crafted URL. No public exploit identified at time of analysis as KEV-listed, but a publicly available proof-of-concept exists on GitHub, slightly elevating practical risk despite the EPSS score of 0.03% (10th percentile).
Stored cross-site scripting in hemant6488's CodeIgniter-StudentManagementSystem allows remote unauthenticated attackers to inject arbitrary JavaScript via the Name argument of the addStudent function in view_students.php. When a victim user views the student listing, the injected script executes in their browser context, enabling session hijacking, credential theft, or defacement. A publicly available proof-of-concept exists via GitHub issue report; however, this vulnerability is not listed in CISA KEV, and EPSS scoring places exploitation probability at 0.03%, indicating low real-world exploitation activity despite POC availability.
Information exposure via verbose SQL error messages in SourceCodester CET Automated Grading System with AI Predictive Analytics 1.0 enables authenticated remote attackers to harvest internal database details by manipulating the /index.php SQL Handler endpoint. The application returns raw SQL error output rather than sanitized application-level messages, leaking schema structure, table names, or query internals. A public proof-of-concept exploit is available on GitHub; this CVE is not listed in the CISA KEV catalog, and the CVSS 4.0 score of 2.1 reflects the low-severity, confidentiality-only impact.
Cross-site scripting in blitz-js blitz (versions 3.0.0-3.0.2) allows remote attackers to inject malicious scripts via the 'Next' redirect parameter in the LoginForm component's Sign-in flow. The vulnerability requires passive user interaction (a victim must follow a crafted link) and is limited to low-integrity impact on the vulnerable system per CVSS 4.0 scoring. Publicly available exploit code exists (GitHub Gist), though EPSS stands at 0.03% (9th percentile) indicating low observed exploitation probability, and it is not listed in CISA KEV. The vendor did not respond to responsible disclosure, meaning no patch has been released.
Reflected cross-site scripting in pingvin-share's sign-in auto-redirect feature allows remote unauthenticated attackers to inject and execute arbitrary JavaScript in a victim's browser by manipulating the `redirect` query parameter. All releases from 1.0 through 1.13.0 (the full release history) are affected. A publicly available proof-of-concept exploit exists on GitHub, and the vendor has not responded to responsible disclosure — meaning no patch has been issued and no vendor advisory exists.
Improper access control in JeecgBoot's AiragModelController (versions up to 3.9.1) permits any authenticated low-privilege user to invoke the list and queryById API endpoints without proper authorization checks, exposing AI RAG model configuration data restricted to higher-privileged roles. The CVSS vector (PR:L, C:L) confirms this is an authorization bypass rather than a full authentication bypass, limiting impact to confidentiality of AI model metadata. Publicly available exploit code exists (GitHub issue #9599, referenced in the exploit tag), though no CISA KEV listing indicates confirmed widespread active exploitation at time of analysis.
SQL injection in CodeAstro Leave Management System 1.0 allows authenticated remote attackers to manipulate the `email_id` parameter in `/admin/add_staff.php`, enabling arbitrary SQL query manipulation against the underlying database. The CVSS 4.0 score of 2.1 (Low) reflects constrained impact scope - confidentiality, integrity, and availability are each rated Low with no lateral impact to subsequent systems - and EPSS at 0.03% (8th percentile) indicates negligible real-world exploitation probability despite a publicly available proof-of-concept on GitHub. No active exploitation has been confirmed and this CVE is not listed in the CISA KEV catalog.
Cross-site request forgery in SourceCodester CET Automated Grading System with AI Predictive Analytics 1.0 enables remote attackers to perform unauthorized state-changing actions by tricking an authenticated user into visiting a malicious page. The CVSS 4.0 vector (VI:L, SC:N/SI:N/SA:N) confirms impact is limited to low-level integrity degradation on the vulnerable system with no confidentiality or availability consequence. A publicly available exploit (PoC HTML page and advisory) has been released to GitHub by researcher NARKHEDE-VAIBHAV; no public exploit identified at time of analysis as confirmed actively exploited (CISA KEV).
Kirby CMS's path resolver fails to enforce `pages.access` permission checks when rendering page drafts, allowing authenticated users who lack access to specific page models to view those drafts by knowing the full URL path. Affected installations are those explicitly configured to restrict certain user roles from accessing pages via `pages.access: false` in user or model blueprints - sites where all users may access all pages are unaffected. No public exploit or CISA KEV listing exists at time of analysis, but the impact is information disclosure of unpublished content such as pre-launch product pages or embargoed posts.
Incorrect permission assignment in the NVIDIA Display Driver kernel module on Windows and Linux allows a highly privileged local user to corrupt critical resource permissions, leading to denial of service and potentially data tampering. Affected product lines span GeForce (limited to Maxwell, Volta, and Pascal architectures on some branches), RTX/Quadro/NVS, Tesla, and Guest (vGPU) drivers across multiple version branches. No public exploit exists and no active exploitation has been identified; SSVC assessment rates exploitation as none and impact as partial, consistent with the moderate CVSS score of 4.4.
Path traversal in XWiki Platform's WebJars API enables a subwiki admin who can publish and install a malicious WebJar extension to write arbitrary files anywhere on the server filesystem. The affected Maven component `xwiki-platform-webjars-api` fails to validate that JAR entry paths extracted during extension installation remain within the intended export directory, allowing overwrite of configuration files or potential superadmin credential manipulation. No public exploit is identified and no CISA KEV listing exists; vendor-released patches are available across three version branches.
Kavita reading server versions prior to 0.9.0 expose seven download and metadata API endpoints without enforcing library-level access controls, enabling authenticated users to retrieve file contents, file sizes, and chapter metadata from libraries they have not been granted access to. The affected endpoints - /api/Download/volume, /api/Download/chapter, /api/Download/series, /api/Chapter, and corresponding size-check variants - accept resource identifiers (chapterId, volumeId, seriesId) without verifying the requesting user's library membership. A proof-of-concept exists per SSVC classification, though EPSS at 0.04% (11th percentile) reflects minimal observed exploitation activity; the CVE is not listed in CISA KEV.
Stored cross-site scripting in SourceCodester Hospitals Patient Records Management System 1.0 allows a remote, high-privileged attacker to inject malicious script via the Remarks argument on the /admin/?page=patients/view_patient endpoint, resulting in low-integrity impact when a victim administrator views the affected patient record. A public proof-of-concept exploit exists (GitHub issue), though the CVSS 4.0 base score of 1.9 and EPSS of 0.03% (8th percentile) indicate low real-world exploitation likelihood. This is not listed in the CISA KEV catalog and SSVC classifies it as non-automatable with partial technical impact.
Path traversal in Joomla! CMS com_media web service endpoint allows authenticated high-privilege attackers to read arbitrary files outside the intended web root via a manipulated search parameter. Affected versions span two major release trains: 4.0.0 through 5.4.5 and 6.0.0 through 6.1.0. No public exploit code exists and no active exploitation has been confirmed, but the vulnerability results in full confidentiality loss of the vulnerable system's filesystem contents accessible to the web process.
Memory leak in GPAC MP4Box up to version 2.4.0 allows a local, low-privileged attacker to exhaust process memory by supplying a crafted MP4 file that triggers the vulnerable `Media_GetSample` function in `src/isomedia/media.c`. The root cause is a missing zero-size guard before memory allocation when the `cat` argument is manipulated to produce a zero-sum of `data_size` and `padding_bytes`. A publicly available proof-of-concept exploit (poc.zip) exists, but the CVSS 4.0 score of 1.9, EPSS of 0.01% (2nd percentile), and strictly local attack vector collectively indicate very low real-world risk; no active exploitation has been identified.
Null pointer dereference in GPAC's MP4Box tool (versions 2.0 through 2.4.0) allows a local, low-privileged attacker to crash the application by supplying a crafted MP4 file with a malformed Protection System Header Box (PSSH). The vulnerability resides in the MergeFragment function, which fails to validate the private_data pointer before passing it to memmove, resulting in a denial-of-service impact limited to the MP4Box process. A public proof-of-concept exploit exists (hosted on GitHub), though EPSS sits at 0.01% (2nd percentile) and the flaw is not listed in the CISA Known Exploited Vulnerabilities catalog, consistent with the low-severity, local-only nature of the issue.
Heap-based buffer overflow in Squirrel versions 3.0 through 3.2 allows a locally authenticated low-privilege attacker to corrupt heap memory by supplying a malicious Cnut file to the ReadObject function in sqobject.cpp. The impact is limited to partial confidentiality, integrity, and availability effects with no scope change, as confirmed by the CVSS 4.0 score of 1.9. A public proof-of-concept exploit exists on GitHub, but this vulnerability has not been confirmed actively exploited by CISA KEV, and EPSS places exploitation probability at just 0.01% (2nd percentile), indicating very low real-world exploitation activity.
Out-of-bounds read in GNU LibreDWG's read_2004_compressed_section function (src/decode.c) affects all versions through 0.14, allowing a local low-privileged attacker to crash the dwgbmp utility or any LibreDWG-based application by supplying a crafted DWG 2004 file with manipulated section address or size fields. Impact is limited to availability (application crash) with no confirmed confidentiality or integrity exposure per the CVSS 4.0 vector. A publicly available proof-of-concept DWG file exists on GitHub, but EPSS at 0.01% (2nd percentile) and no CISA KEV listing confirm this is not currently subject to widespread exploitation.
Null pointer dereference in GNU LibreDWG's dwggrep utility crashes the application when processing a maliciously crafted DWG file. The vulnerability resides in the match_BLOCK_HEADER function within dwggrep.c and affects all tracked releases from version 0.1 through 0.14. A local authenticated attacker can exploit this to cause denial of service against the dwggrep utility; no publicly available exploit code exists for confidentiality or integrity compromise, consistent with the CVSS impact scores of VC:N/VI:N/VA:L. Publicly available exploit code exists (no KEV listing), though EPSS at 0.01% reflects negligible widespread exploitation probability.
VM escape in Kata Containers allows any Kubernetes user with pod-creation rights to break out of the VM sandbox and gain full read/write access to the host filesystem. All Kata Containers installations prior to commit ffa59ce3aa78 are affected when using the default configuration.toml, which enables the `virtio_fs_extra_args` and `kernel_params` pod annotations out of the box. An attacker crafts a pod with two annotations: one to redirect virtiofsd to serve the host root filesystem (`/`) into the guest VM, and a second to enable the agent debug console - after which the entire host filesystem is accessible from inside the supposedly isolated VM. A fully working proof-of-concept with confirmed output against Kata Containers 3.28.0 on Ubuntu 24.04 has been publicly disclosed; no public exploit confirmed as actively exploited (CISA KEV) at time of analysis.