LibreNMS before 24.10.0 allows a remote attacker to execute arbitrary code via OS command injection involving AboutController.php's index(), SettingsController.php's update(), and PollDevice.php's. Rated critical severity (CVSS 9.1), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and EPSS exploitation probability 60.2%.
SQL injection in LiteLLM proxy server versions 1.81.16 through 1.83.6 allows unauthenticated remote attackers to read and modify database contents, gaining unauthorized access to managed LLM API credentials. The vulnerability is exploitable via crafted Authorization headers sent to any LLM API route (e.g., POST /chat/completions), triggering the injection through the proxy's error-handling path. Vendor-released patch available in version 1.83.7. No active exploitation confirmed (not in CISA KEV), but the attack vector is simple (CVSS 4.0: AV:N/AC:L/PR:N) and SQL injection POCs are widely known. Discovered by Tencent YunDing Security Lab.
Remote code execution in dash-uploader (Python package for Plotly Dash) versions 0.1.0 through 0.7.0a2 allows unauthenticated remote attackers to execute arbitrary code via directory traversal flaws in the HTTP request handler. The vulnerability affects temp_root path handling and POST request processing, enabling attackers to write files outside intended upload directories. Public exploit code exists (GitHub repository CVE-2026-38360), and the CVSS 9.8 critical score reflects the network-accessible, no-authentication attack vector. EPSS data not available, but the combination of RCE impact, public POC, and trivial exploitation complexity (AC:L/PR:N) makes this a high-priority remediation target for any deployment using vulnerable dash-uploader versions.
Certain GL.iNet devices with 4.x firmware allow authentication bypass (resulting in administrative control of the device) via a username that is both a valid SQL statement and a valid regular. Rated critical severity (CVSS 9.8), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Public exploit code available and no vendor patch available.
Arbitrary local code execution in electerm (versions 3.0.6-3.8.14) allows remote attackers to execute malicious code on victim systems by tricking users into clicking crafted electerm:// deep links, opening malicious shortcuts, or running CLI commands with attacker-controlled --opts parameters. The vulnerability stems from insufficient input validation (CWE-20) on deep link and CLI arguments, enabling adversaries to inject arbitrary options that execute code with the privileges of the electerm process. Exploitation requires user interaction (clicking link or opening file) but no authentication, making it suitable for phishing or watering-hole attacks. Patch available in version 3.8.15 with deny-list controls blocking critical parameter override.
Remote authentication bypass in Open WebUI LDAP integration (versions ≤0.8.12) allows complete account takeover by submitting empty passwords. The vulnerability exploits RFC 4513 unauthenticated simple bind semantics: when LDAP is enabled, attackers can authenticate as any user-including administrators-with zero knowledge of actual passwords, gaining full access to chats, files, API keys, and settings. Affects deployments using OpenLDAP default configurations or certain Active Directory setups that accept empty-password binds. Vendor-released patch: version 0.9.0. CVSS 9.1 (Critical) reflects network-accessible, zero-privilege, zero-interaction exploitation with high confidentiality and integrity impact.
Remote code execution as root in Remote Spark SparkView before build 1122 allows network attackers to bypass local connection authentication checks and execute arbitrary commands with maximum privileges. CVSS 4.0 assigns the maximum 10.0 score with network vector, low complexity, and no authentication required (AV:N/AC:L/PR:N). The vendor description explicitly warns that depending on implementation, unauthenticated attackers can exploit this flaw. EPSS and KEV data not provided, but the combination of trivial exploitation conditions and root-level impact makes this critical for any organization running affected SparkView builds.
Remote unauthenticated command injection in Universal Robots PolyScope Dashboard Server (versions <5.21.1) allows attackers to execute arbitrary OS commands on industrial robot controllers via network-crafted requests. With CVSS 9.8 (critical severity) and complete absence of authentication barriers, this vulnerability enables full robot controller compromise from remote network positions. No authentication, user interaction, or attack complexity required - exploitation is straightforward against default configurations exposing the Dashboard Server interface.
Unauthorized API access in sovity Dataspace Portal versions 2.1.1 through 7.3.1 allows unauthenticated remote attackers to bypass authorization controls and access backend APIs using credentials from self-registered accounts in PENDING status. The vulnerability affects the open-source SaaS platform before organizations approve new user registrations, enabling information disclosure and potential data manipulation. Vendor-confirmed patch released in version 7.3.2 on 2026-04-20. CVSS 10.0 reflects network-accessible attack with no complexity, no privileges required, and high impact across confidentiality, integrity, and availability in both vulnerable and subsequent systems. No CISA KEV listing or public exploit identified at time of analysis.
Missing authentication on the free5GC Session Management Function (SMF) lets any network attacker reaching the SBI read, inject, and delete User-Plane topology (UP-node/link) data without credentials. In free5GC through 4.2.1 the UPI management route group is mounted without the OAuth2/bearer-token middleware that its sibling nsmf-oam group has, so unauthenticated GET/POST/DELETE requests on /upi/v1/upNodesLinks reach the business handlers. A researcher-demonstrated PoC exists (validated 2026-03-13 against the official Docker lab); the CVE is not in CISA KEV and EPSS is very low (0.05%, 15th percentile), indicating no observed widespread exploitation despite the CVSS 10.0 score.
Authentication bypass in free5GC's NEF (Network Exposure Function) nnef-pfdmanagement SBI route group allows any network attacker reaching the NEF service interface to submit forged bearer tokens and access PFD (Packet Flow Description) management APIs on free5GC versions before 4.2.2. Because the entire route group is mounted with no inbound OAuth2/authorization middleware despite NEF advertising 'OAuth2 setting receive from NRF: true', attackers can read AF-supplied traffic-classification policy, create attacker-controlled change-notification subscriptions (an outbound SSRF primitive), and delete legitimate subscriptions. Publicly available exploit code exists with an end-to-end validated PoC, though EPSS remains very low (0.04%) and it is not listed in CISA KEV.
Authentication bypass in free5GC's NEF (Network Exposure Function) lets any host that can reach the Service-Based Interface (SBI) call the nnef-oam route group with no Authorization header and receive HTTP 200, despite the NEF advertising 'OAuth2 setting receive from NRF: true'. The defect is route-group-scoped (CWE-306): the OAM group is mounted with no inbound auth middleware, so the current stub (which returns null) and any future OAM read/modify/restart operation added to the group inherit an open, unauthenticated boundary. A working PoC (a single unauthenticated curl) is published in the GHSA advisory; EPSS is very low (0.04%) and it is not in CISA KEV.
SQL injection in Emlog's article creation and update functions allows attackers to inject arbitrary SQL through unsanitized input, enabling full database compromise, data theft, or destruction of stored content in all versions prior to 2.6.11. The flaw carries a vendor CVSS 4.0 base score of 10.0 and CWE-89 classification; publicly available exploit code exists (SSVC exploitation status: poc) but there is no evidence of active in-the-wild abuse. EPSS is very low (0.03%), consistent with an obscure, low-footprint CMS rather than a mass-targeted enterprise product.
Remote code execution in ai-scanner versions 1.0.0 through 1.4.0 allows authenticated attackers to inject and execute arbitrary JavaScript code via the BrowserAutomation::PlaywrightService component. The vulnerability has a Critical CVSS score of 9.9 with scope change, enabling cross-boundary compromise of confidentiality, integrity, and availability. Vendor-released patch available in version 1.4.1 as of April 13, 2026, with GitHub Security Advisory GHSA-r27j-xxgx-f5vr confirming the fix.
Remote code execution in Termix (self-hosted web-based server management platform) before version 2.1.0 lets an authenticated user run arbitrary OS commands on any managed remote server. The Docker container-management endpoints pass the attacker-controlled containerId (from the URL path and WebSocket messages) straight into shell commands run over SSH via ssh2.Client.exec() with no sanitization, so a crafted container ID achieves full command injection. Rated CVSS 9.9 with no public exploit identified at time of analysis; EPSS is very low (0.07%) and it is not on CISA KEV.
Authenticated code execution in Netgate pfSense CE 2.8.0 is claimed via the XMLRPC API method pfsense.exec_php, which runs attacker-supplied PHP on the firewall. Netgate publicly disputes this as a vulnerability because pfsense.exec_php is restricted to administrators, who are by design permitted to execute PHP, meaning the CVSS 9.9 rating reflects an intended admin capability rather than a privilege boundary being crossed. Publicly available exploit code exists (SSVC PoC), but EPSS is 0.01% and it is not in CISA KEV, indicating no observed active exploitation.
Remote code execution in FastGPT's agent-sandbox component (versions 4.14.10 through 4.14.12) lets any network-adjacent, unauthenticated attacker seize full control of the sandbox environment. The bundled entrypoint.sh launches code-server with --auth none and binds it to 0.0.0.0:8080, so reaching the exposed port grants an interactive VS Code-based shell with no password. CVSS is 9.8 and SSVC rates technical impact as total with automatable exploitation; publicly available exploit code exists (SSVC 'poc'), though EPSS remains modest at 0.33%.
Command injection in electerm's npm install script allows arbitrary command execution on macOS systems during 'npm install -g electerm'. The runMac() function in install.js:150 passes attacker-controlled remote release metadata (releaseInfo.name) directly to exec('open ...') without validation, enabling remote code execution as the installing user. CVSS 9.8 (AV:N/AC:L/PR:N/UI:N) reflects theoretical network-based exploitation, though actual attack requires compromise of the project's update server or man-in-the-middle position during npm package installation. No public exploit identified at time of analysis. Vendor-released patch: version 3.3.8 (commit 59708b3).
Arbitrary code execution and privileged token theft affects Postiz (gitroomhq/postiz-app) prior to commit da44801 through an insecure GitHub Actions 'Pwn Request' pattern in the Build and Publish PR Docker Image workflow. Because the workflow triggers on pull_request_target with 'permissions: write-all' yet checks out and builds the untrusted fork head (Dockerfile.dev), any unauthenticated GitHub user can open a fork PR with a malicious Dockerfile.dev to run code on the runner and exfiltrate a write-all GITHUB_TOKEN, effectively hijacking the repository's CI/CD supply chain. Rated CVSS 10.0 with SSVC technical impact 'total' and automatable 'yes', but EPSS is low (0.14%) with no public exploit or KEV listing at time of analysis.
Remote code execution in vm2 npm package (versions ≤3.11.1) allows attackers to escape the JavaScript sandbox via a prototype pollution technique targeting the neutralizeArraySpeciesBatch method. By installing a setter on Array.prototype[0] and triggering Buffer allocation, attackers gain access to the host Function constructor and can execute arbitrary system commands. Publicly available proof-of-concept exists (GHSA-9qj6-qjgg-37qq). CVSS 9.8 with network vector reflects the risk when vm2 is used to execute untrusted code in server-side applications. Vendor-released patch: vm2 v3.11.2 addresses this and two other concurrent sandbox escapes.
Remote code execution in math-codegen npm package versions prior to 0.4.3 allows unauthenticated attackers to execute arbitrary system commands via string literal injection into the cg.parse() function. The vulnerability stems from unsanitized string literals being injected directly into new Function() bodies, enabling full command execution on any application exposing math evaluation endpoints that process user input. EPSS score not available, but this is a critical unauthenticated RCE requiring no special conditions beyond user input reaching the vulnerable parser. Vendor-released patch available in version 0.4.3.
Remote code execution in VM2 (npm package) allows complete sandbox escape via null-prototype exception handling flaw. Attackers can execute arbitrary system commands on the host by exploiting a logic error in the exception proxy mechanism that incorrectly handles objects with null prototypes. Public exploit code exists and the vulnerability affects all versions prior to 3.11.2. The CVSS 9.8 severity reflects network-accessible, unauthenticated exploitation requiring no user interaction - however, real-world risk depends on whether untrusted users can supply code to the VM2 sandbox in a given deployment.
NornicDB's Bolt server binds to all network interfaces (0.0.0.0) regardless of the --address CLI flag or server.host configuration, exposing the graph database with default admin:password credentials to any device on the same LAN. The HTTP server correctly honors bind address restrictions, but a configuration plumbing bug prevents the Bolt protocol listener from reading the intended host parameter. Vendor-released patch available in version 1.0.42-hotfix addresses the underlying CWE-1392 (Improper Binding of Resource to Another Sphere) by adding Host field to Bolt configuration and wiring the resolveBindAddress() function to both protocol listeners. GitHub security advisory GHSA-2hp7-65r3-wv54 confirms the vulnerability with reproduction steps showing netstat evidence of wildcard binding despite localhost configuration.
Use-after-free in the Linux kernel's ksmbd SMB server (smb2_open()) allows remote attackers to potentially trigger memory corruption when accessing an opinfo pointer dereferenced after rcu_read_unlock(). The flaw is fixed in upstream stable releases (6.1.167, 6.6.130, 6.12.78, 6.18.19, 6.19.9, and 7.0); no public exploit identified at time of analysis, and EPSS exploitation probability is very low at 0.02%.
Integer overflow in Linux kernel's IPv6 IOAM (In-situ Operations, Administration, and Maintenance) trace functionality allows remote unauthenticated attackers to trigger buffer overflow conditions. A crafted IOAM trace packet with specific schema configurations causes an 8-bit integer wraparound that bypasses buffer boundary checks, enabling memory corruption with potential for arbitrary code execution at kernel privilege level. CVSS scored 9.8 (Critical) with network attack vector, though EPSS score of 0.02% (7th percentile) suggests minimal observed exploitation activity. Patches available across multiple stable kernel versions (5.15, 6.1.168, 6.6.134, 6.12.81, 6.18.22, 6.19.12, 7.0) via upstream commits, indicating vendor-prioritized remediation without confirmed active exploitation.
Improper key length validation in the Linux kernel's libceph authentication subsystem allows remote unauthenticated attackers to trigger memory corruption during Ceph authentication key decoding. This affects systems using Ceph distributed storage clusters, with EPSS probability 0.02% (percentile 7%), indicating low immediate exploitation likelihood. Patches available across all supported kernel branches (5.15.202, 6.1.165, 6.6.128, 6.12.75, 6.18.16, 6.19.6, 7.0) with commits linked in multiple stable trees, suggesting coordinated vendor response. No public exploit code or CISA KEV listing identified at time of analysis.
Code injection in 1C-Bitrix (Bitrix Site Manager) through 25.100.500 lets an actor holding SOURCE/WRITE permissions on the Translate Module upload a PHP file together with a .htaccess file and have it executed by the web server, yielding arbitrary code execution on the host. Publicly available exploit code exists (karmainsecurity PoC), but there is no public exploit identified as actively used in attacks, and EPSS is very low at 0.02%. Notably, the vendor disputes this as a vulnerability, calling it intended behavior for high-privileged users who legitimately upload translated pages.
Double-free vulnerability in Linux kernel's qla2xxx SCSI driver allows remote code execution or denial of service through malformed Fibre Channel ELS commands. The qla24xx_els_dcmd_iocb() function incorrectly frees fcport structures twice during error handling - once via kref_put() calling qla2x00_els_dcmd_sp_free(), then again explicitly afterward. Despite the CVSS:3.1/AV:N score of 9.8, the network vector appears to reflect the driver's network-facing nature (Fibre Channel over IP or similar) rather than internet-accessible exploitation. EPSS score of 0.02% (5th percentile) indicates extremely low observed exploitation probability. Patches available across multiple stable kernel branches (6.9+, 6.19.9+, 7.0+) per upstream commits.
Timing attacks can compromise TCP Authentication Option (TCP-AO) message authentication codes in Linux kernel 6.7+ due to non-constant-time MAC comparison. Remote unauthenticated attackers on the network path can exploit timing differences during MAC validation to extract authentication secrets, defeating TCP-AO's connection authentication mechanism. Exploitation probability is low (EPSS 0.02%, 5th percentile) with no confirmed active exploitation, but vendor patches are available for affected stable branches including 6.12.78, 6.18.19, 6.19.9, and mainline 7.0.
Use-after-free in Linux kernel ksmbd allows remote unauthenticated attackers to potentially execute arbitrary code, disclose sensitive information, or cause denial of service. The vulnerability stems from improper RCU lock handling in smb_lazy_parent_lease_break_close() where opinfo pointer is dereferenced after RCU read unlock, creating a race condition. Patches available across multiple kernel versions (6.6.130, 6.12.78, 6.18.19, 6.19.9, 7.0). Despite critical CVSS 9.8 score, EPSS exploitation probability is low (0.02%, 5th percentile) and no active exploitation or public POC identified at time of analysis.
Use-after-free in Linux kernel ksmbd allows remote unauthenticated attackers to execute arbitrary code, escalate privileges, or cause denial of service by racing oplock_info access during concurrent RCU read operations. The vulnerability stems from immediate kfree() without RCU grace period, enabling opinfo_get() to call atomic_inc_not_zero() on freed memory. CVSS 9.8 reflects network exploitability without authentication, though EPSS score of 0.02% (5th percentile) suggests minimal observed exploitation attempts. Vendor patches available across multiple kernel versions (6.6.130, 6.12.78, 6.18.19, 6.19.9, 7.0) with fixes referenced in five upstream commits. Not listed in CISA KEV; no public exploit code identified at time of analysis.
Local privilege escalation in Raynet RayVentory Scan Engine through version 12.6 Update 8 stems from an uncontrolled search path (CWE-427): if an attacker can set the PATH environment variable used by the process, they can force execution of a malicious binary and gain elevated privileges. Publicly available exploit code exists (GitHub, Wise-Security), and CISA's SSVC framework marks technical impact as total and the flaw as automatable, but there is no confirmed active exploitation and EPSS is negligible (0.02%). Note the CVE is officially DISPUTED, since controlling the PATH environment is characterized as a site-specific misconfiguration rather than an inherent product flaw.
Reference counting flaw in mlx5e network driver causes kernel memory corruption when XDP multi-buffer programs modify packet layouts via bpf_xdp_pull_data() or bpf_xdp_adjust_tail(). Affects Linux kernel versions 6.18.x through 6.19.9, with vendor patches available for 6.18.19, 6.19.9, and 7.0. The vulnerability triggers negative page pool reference counts leading to memory management errors, discovered by the drivers/net/xdp.py selftest. While CVSS scores this 9.8 Critical with network vector, the technical context suggests local impact requiring specific XDP program execution. EPSS exploitation probability is low (0.02%, 4th percentile) with no evidence of active exploitation or public POC at time of analysis.
Use-after-free in Linux kernel kthread subsystem enables memory corruption leading to arbitrary code execution or denial of service. The vulnerability arises when kernel threads exit via make_task_dead() instead of kthread_exit(), bypassing affinity_node cleanup. This causes dangling pointers in the global kthread_affinity_list that corrupt freed memory reused by the SLAB allocator, specifically overwriting RCU callback function pointers in struct pid objects. CVSS rates this 9.8 critical, though the network attack vector appears misclassified since kernel thread manipulation requires local code execution. EPSS score of 0.02% (4th percentile) indicates low predicted exploitation likelihood despite severity. Vendor patches available for Linux 6.18.19, 6.19.9, and 7.0 via upstream commits.
SQL injection in Beauty Parlour Management System v1.1 enables unauthenticated remote attackers to extract, modify, or delete database contents through the aptnumber parameter at /appointment-detail.php endpoint. With CVSS 9.8 (critical severity) and network-accessible exploit requiring no authentication, this represents a complete compromise vector. Public proof-of-concept code exists on GitHub, and CISA SSVC framework rates it as automatable with total technical impact, though CISA KEV does not yet list active exploitation. EPSS data unavailable, but the combination of public POC, zero authentication requirements, and straightforward SQLi exploitation pattern indicates high probability of opportunistic scanning and exploitation.
Command injection in PraisonAI's MCP server command handler enables remote unauthenticated attackers to execute arbitrary operating system commands. The vulnerability exists in parse_mcp_command() which accepts MCP server commands without validating executables or arguments, allowing injection of shell commands like 'bash -c', 'python -c', or '/bin/sh -c' with inline code execution. GitHub security advisory GHSA-9qhq-v63v-fv3j confirms this is an incomplete fix for CVE-2026-34935. Vendor-released patch version 4.6.9 (upstream version 1.5.69) implements an allowlist of permitted MCP executables and validates commands against ALLOWED_MCP_COMMANDS. No active exploitation confirmed (not in CISA KEV); proof-of-concept exploit code published in advisory demonstrates trivial exploitation.
Arbitrary code execution in Electerm terminal client (≤3.8.15) allows attackers who control terminal output to execute commands or access local files when victims click hyperlinks. The unvalidated shell.openExternal call accepts any protocol scheme, enabling 'file://' URIs for local file access or platform-specific handlers for code execution. No vendor-released patch identified at time of analysis. GitHub Security Advisory GHSA-fwf6-j56g-m97c confirms the vulnerability. CVSS 9.6 reflects high impact across confidentiality, integrity, and availability with scope change, though exploitation requires user interaction (clicking a malicious link).
Cross-origin WebSocket hijacking in the Cline `kanban` server (npm `kanban`, shipped via the `cline` CLI ≤ 2.13.0) lets any website a developer visits silently connect to the loopback server on 127.0.0.1:3484 and leak workspace paths, task content, git branches, and AI-agent chat, then inject prompts into a running agent's terminal for remote code execution and kill active tasks. Because browsers do not apply CORS to WebSocket handshakes and the three endpoints perform no Origin validation or authentication (CWE-306), exploitation needs only that the victim run kanban and browse to an attacker page. A full interactive proof-of-concept exists (publicly available exploit code exists); it is not listed in CISA KEV and EPSS is very low at 0.02%, indicating no observed mass exploitation to date.
Remote code execution in SiYuan's Electron renderer occurs when users hover over search results, file tree items, or attribute view elements containing URL-encoded XSS payloads in document titles or metadata. The vulnerability chains a URL-decoding step (decodeURIComponent) with unsafe innerHTML assignment in tooltip rendering, bypassing the escapeAriaLabel sanitizer that only handles HTML entities but ignores %XX URL escapes. Because SiYuan's renderer runs with nodeIntegration:true and contextIsolation:false, the XSS escalates to arbitrary code execution via require('child_process'). Exploitation requires user interaction (hovering) but no authentication, and malicious payloads survive .sy.zip export/import and sync replication, enabling supply-chain and shared-workspace attacks. No public exploit code identified at time of analysis, though detailed proof-of-concept is published in the GitHub advisory.
Remote code execution in SiYuan's Electron desktop application allows authenticated attackers (or browser extensions on localhost) to inject malicious JavaScript through unescaped Attribute View names, escalating from stored XSS to arbitrary system command execution. The Go kernel backend stores AV names without HTML escaping, then embeds them via string replacement into HTML templates pushed over WebSocket. Three TypeScript renderer paths (render.ts, Title.ts, transaction.ts) consume this data using innerHTML/outerHTML without sanitization. Because the Electron main window runs with nodeIntegration:true and contextIsolation:false, script injection grants full Node.js API access—enabling attackers to spawn child processes (calc.exe/xcalc demonstrated in PoC), exfiltrate SSH keys, install backdoors, or pivot to cloud credentials. Payloads persist in JSON files under data/storage/av/, replicate across all sync transports (S3/WebDAV/cloud), survive .sy.zip export-import, and trigger for any user role (Administrator/Editor/Reader/Visitor) opening a document bound to the poisoned database view. CVSS 9.4 (Network/Low/None/High Confidentiality-Integrity-Availability + Scope Changed) reflects worst-case remote network vector, though the primary realistic attack path is via installed browser extensions (chrome-extension:// Origin explicitly allowlisted in session.go:277) calling the /api/transactions endpoint as an auto-granted admin on default installations with no Access Authorization Code. GitHub advisory GHSA-2h64-c999-c9r6 confirms patch available in kernel commit 0.0.0-20260512140701-d7b77d945e0d. No public exploit code identified at time of analysis, but detailed reproduction steps with curl payloads and Electron DevTools inspection are published in the advisory.
Arbitrary file write in PraisonAI's MCP server escalates to remote code execution through path traversal when user interaction triggers malicious tool calls. The praisonai mcp serve daemon accepts attacker-controlled path arguments without validation, allowing writes outside the intended ~/.praison/rules/ directory. Attackers can drop Python .pth files into site-packages to achieve code execution in any subsequent Python process run by the victim user. CVSS 9.4 with network vector and low complexity, though exploitation requires user interaction (PR:N/UI:P). No active exploitation confirmed (not in CISA KEV) and no public POC identified at time of analysis, but the detailed advisory provides sufficient information for weaponization.
Unauthenticated CRUD access to free5GC's NEF 3gpp-traffic-influence API (v4.2.1, image v4.2.0) lets any host that can reach the NEF Service-Based Interface create, read, patch, and delete traffic-influence subscriptions with no Authorization header or a forged bearer token. Attackers can inject AnyUeInd=true group/any-UE subscriptions to redirect subscriber traffic to attacker-chosen DNAIs/routing endpoints via SMF/UPF, leak steering policy, or delete legitimate subscriptions. Publicly available exploit code exists (full end-to-end PoC in the advisory); EPSS is low (0.04%, 11th percentile) and there is no public exploit identified as actively exploited.
Authentication bypass in free5GC's NEF (Network Exposure Function) through v4.2.1 lets any party with SBI network reachability create, read, and delete 3gpp-pfd-management transaction state using a forged or arbitrary bearer token such as 'Bearer not-a-real-token'. Because the PFD-management route group is mounted with no inbound OAuth2 middleware - and remains reachable even when the runtime ServiceList omits it - attackers can poison PFD policy written to UDR (affecting downstream SMF/UPF traffic classification), leak AF-supplied policy data, or delete provisioned rules. Publicly available exploit code exists (full end-to-end PoC in the GHSA advisory), though EPSS is low (0.04%, 11th percentile) and it is not in CISA KEV.
Timing attacks against TCP MD5 authentication in Linux kernel allow remote attackers to forge connection signatures through MAC comparison oracle. The vulnerability exists because MAC (Message Authentication Code) comparisons in the TCP-MD5 implementation are not constant-time, enabling attackers to extract authentication secrets through timing side-channels. All Linux kernel versions from 2.6.20 through 6.19.9 are affected. Patches are available across all actively maintained stable branches (5.10, 6.1, 6.6, 6.12, 6.18, 6.19, 7.0). EPSS score of 0.02% suggests low automated exploitation probability, though the network-accessible attack vector and authentication bypass capability represent significant risk for systems using TCP MD5 signatures (RFC 2385).
Denial of service in Xen's oxenstored (the OCaml Xenstore daemon) arises because quota-related use counts are not released when a domain is destroyed, per Xen Security Advisory 483 (XSA-483). A malicious or buggy guest can repeatedly create and destroy Xenstore state so that leaked accounting counters permanently consume quota, eventually preventing legitimate Xenstore operations and denying service to the host control plane and other domains. There is no public exploit identified at time of analysis, and the issue is not listed in CISA KEV.
Local privilege escalation in the Xen hypervisor, disclosed as one of five issues under Xen Security Advisory XSA-489 (alongside CVE-2026-23559/23560/23562/42486), lets a malicious or compromised guest run code with more privilege than intended and cross the guest/host trust boundary. The CVSS 4.0 vector (AV:L, scope-changed with high confidentiality, integrity and availability impact to subsequent systems) indicates a guest can affect the underlying host and other guests, consistent with the CWE-250 'execution with unnecessary privileges' root cause. No public exploit identified at time of analysis, and it is not listed in CISA KEV.
Privilege escalation in the Xen hypervisor (addressed in Xen Security Advisory XSA-489) allows an attacker executing within a guest VM to break isolation and compromise the underlying host, per its CVSS 4.0 subsequent-system impact of High across confidentiality, integrity, and availability. The flaw is rooted in execution with unnecessary privileges (CWE-250) and is one of several issues (CVE-2026-23559 through -23562 and CVE-2026-42486) disclosed together in XSA-489 v2. There is no public exploit identified at time of analysis and it is not listed in CISA KEV, but a CVSS 4.0 base score of 9.4 marks it as critical for multi-tenant virtualization environments.
Local privilege escalation in the Xen hypervisor, disclosed as part of Xen Security Advisory 489 (XSA-489, which bundles CVE-2026-23559 through CVE-2026-23562 and CVE-2026-42486), allows a malicious or compromised guest domain to break guest/host isolation and gain full control over the hypervisor and co-tenant guests. The CWE-250 (Execution with Unnecessary Privileges) root cause combined with the CVSS 4.0 scope-changed, high-impact-across-the-board vector (VC:H/VI:H/VA:H/SC:H/SI:H/SA:H) indicates a guest-to-host escalation rather than a mere in-guest issue. There is no public exploit identified at time of analysis, and the vulnerability is not listed in CISA KEV.
Privilege escalation in the Xen hypervisor (disclosed as XSA-489, CVE-2026-23559, one of a bundle of RB-tree/related flaws alongside CVE-2026-23560/23561/23562/42486) allows a malicious or compromised guest to escape guest confinement and compromise the host, with full confidentiality, integrity and availability impact on both the guest and the surrounding system. The CVSS 4.0 vector (9.4) indicates a local attack surface with scope-changed high subsequent-system impact, consistent with a guest-to-hypervisor escalation. There is no public exploit identified at time of analysis and it is not listed in CISA KEV; disclosure is currently pre-NVD via the oss-security mailing list.
Privilege escalation in the Xen hypervisor (tracked in Xen Security Advisory XSA-489, bundled with CVE-2026-23559 through CVE-2026-23562) allows a local attacker operating within a guest context to break isolation and gain control over the underlying host. The CWE-250 root cause (execution with unnecessary privileges) combined with a scope-changing CVSS 4.0 vector (9.4) means a compromised or malicious guest can impact the host and other co-resident guests. This was disclosed pre-NVD via the oss-security mailing list on 2026/04/29; no public exploit identified at time of analysis and it is not listed in CISA KEV.