Debian

Remote denial-of-service in vouch-proxy v0.47.2 and earlier crashes the server process with a single unauthenticated HTTP request. The /validate endpoint's multipart cookie reassembly routine in pkg/cookie/cookie.go parses the total part count directly from the attacker-controlled Cookie header name - an N value of 10 billion passed as VouchCookie_1of10000000000 triggers a make([]string, N) call requesting approximately 160 GB of heap, immediately raising a Go runtime fatal out-of-memory error. No authentication, session token, or prior interaction is required; a fully working PoC with Docker reproduction artifacts is publicly disclosed in the GitHub Security Advisory, though no CISA KEV listing is confirmed at time of analysis.

Arbitrary root file read in Linuxfabrik monitoring plugins allows a local attacker controlling the nagios or icinga account to exfiltrate any root-readable file by exploiting the hidden but permanently live --test argument accepted by all sudoers-whitelisted check plugins. Invoking the deb-updates plugin via sudo with --test=/etc/shadow,,0 causes lftest.py to read the supplied path as root with no path confinement and echo its full contents to stdout; approximately 22 additional whitelisted plugins yield filtered leaks or file-existence oracles through the same mechanism. No active exploitation has been confirmed by CISA KEV, but a fully functional proof-of-concept command is publicly disclosed in the vendor's own GitHub security advisory GHSA-rh9c-rqvg-f7pr, and patches are available in linuxfabrik-lib v6.1.0 and monitoring-plugins v7.0.0.

Local privilege-escalation/denial-of-service via a use-after-free in the Linux kernel's KVM x86 I/O APIC emulation (CVE-2026-74517): during VM teardown the delayed I/O APIC EOI work (kvm_ioapic_eoi_inject_work) can execute after the target vCPUs are freed, causing __kvm_irq_delivery_to_apic_fast to read freed slab memory. Any local principal able to create and run a KVM guest can race VM destruction against a pending delayed EOI to corrupt host kernel memory. EPSS is low (0.15%, 5th percentile) and the issue is not on CISA KEV; no public exploit is identified, and the reporter/tags frame the practical impact as denial of service.

Invalid pointer dereference in the Linux kernel netfilter NAT subsystem exposes local attackers with CAP_NET_ADMIN to kernel slab-out-of-bounds memory access via nf_nat_register_fn(). When nf_nat_init() fails during module initialization, the nat_net pointer is left in an invalid state; subsequent calls to nf_nat_register_fn() dereference it without validation, producing a KASAN-confirmed out-of-bounds read that can escalate to full kernel memory corruption. No public exploit code exists and EPSS at 0.21% (12th percentile) reflects negligible observed exploitation activity, though the local-privilege attack surface in containerized environments warrants timely patching.

Stack-out-of-bounds read in the Linux kernel's XFRM/IPsec subsystem allows a local user with network policy configuration privileges to read out-of-bounds kernel stack memory and potentially crash the system by crafting an outbound XFRM policy using an optional IPTFS-mode template. The flaw is an incomplete fix: commit 3d776e31c841 blocked optional tunnel and BEET templates in outbound policies to prevent this exact class of bug, but the later-added IPTFS mode was never covered by that restriction. No active exploitation is confirmed (not in CISA KEV), EPSS is 0.15%, and a functional reproducer is embedded in the upstream fix commit, lowering the barrier for any local attacker who discovers the configuration.

Use-after-free race condition in the Linux kernel's TC act_tunnel_key subsystem allows a local low-privileged attacker to corrupt kernel heap memory, potentially escalating privileges or crashing the system. The flaw exists in tunnel_key_release_params(), which synchronously frees metadata_dst via dst_release() while concurrent RCU readers in tunnel_key_act() may still hold the old params pointer and attempt dst_clone() on the already-dead reference. A proof-of-concept was developed by Trend Micro ZDI (zdi-disclosures@trendmicro.com) and verified by kernel maintainers with KASAN confirming slab-use-after-free; no public exploit is available beyond that researcher-held POC, and no KEV listing exists at time of analysis.

In the Linux kernel, the following vulnerability has been resolved: KVM: VMX: Grab vmcs12 on CR8 interception update iff vCPU is in guest mode When updating CR8 intercepts, get vmcs12 if and only if the vCPU is in guest mode so that a future change can have update CR8 intercepts during vCPU creation, without running afoul of get_vmcs12()'s lockdep assertion. ------------[ cut here ]------------ debug_locks && !(lock_is_held(&(&vcpu->mutex)->dep_map) || !refcount_read(&vcpu->kvm->users_count)) WARNING: arch/x86/kvm/vmx/nested.h:61 at get_vmcs12 arch/x86/kvm/vmx/nested.h:60 [inline], CPU#0: syz.2.19/5879 WARNING: arch/x86/kvm/vmx/nested.h:61 at vmx_update_cr8_intercept+0x3de/0x4e0 arch/x86/kvm/vmx/vmx.c:6879, CPU#0: syz.2.19/5879 Modules linked in: CPU: 0 UID: 0 PID: 5879 Comm: syz.2.19 Not tainted syzkaller #0 PREEMPT(full) Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.2-debian-1.16.2-1 04/01/2014 RIP: 0010:get_vmcs12 arch/x86/kvm/vmx/nested.h:60 [inline] RIP: 0010:vmx_update_cr8_intercept+0x3de/0x4e0 arch/x86/kvm/vmx/vmx.c:6879 Call Trace: <TASK> apic_update_ppr arch/x86/kvm/lapic.c:984 [inline] kvm_lapic_reset+0x1c24/0x2980 arch/x86/kvm/lapic.c:3023 kvm_vcpu_reset+0x44c/0x1bf0 arch/x86/kvm/x86.c:12986 kvm_arch_vcpu_create+0x746/0x8b0 arch/x86/kvm/x86.c:12847 kvm_vm_ioctl_create_vcpu+0x428/0x930 virt/kvm/kvm_main.c:4201 kvm_vm_ioctl+0x893/0xd50 virt/kvm/kvm_main.c:5159 vfs_ioctl fs/ioctl.c:51 [inline] __do_sys_ioctl fs/ioctl.c:597 [inline] __se_sys_ioctl+0xfc/0x170 fs/ioctl.c:583 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline] do_syscall_64+0x174/0x580 arch/x86/entry/syscall_64.c:94 entry_SYSCALL_64_after_hwframe+0x77/0x7f </TASK> No functional change intended.

Heap buffer overflow in libXfont2's font-server client allows a malicious or compromised font server to overflow the X server's heap with attacker-controlled content by sending overlapping glyph source offsets. Affected systems running X server as root are exposed to privilege escalation; unprivileged X server instances crash with denial of service. All libXfont2 versions prior to 2.0.9 are affected. No public exploit code or CISA KEV listing has been identified at time of analysis, but the attacker-controlled write primitive makes this a high-severity memory corruption issue.

Out-of-bounds heap read and write in libXfont2's font server client (all versions prior to 2.0.9) allows a malicious or compromised network font server to corrupt X server heap memory, enabling privilege escalation when the X server runs as root or denial of service via crash in unprivileged configurations. The flaw arises because fs_read_glyphs() in src/fc/fserve.c allocates the encoding[] array based on num_extents from one protocol reply but indexes it using num_chars from a separate reply with no cross-validation, so a crafted server can send mismatched values to trigger OOB heap reads and attacker-controlled OOB heap writes. No public exploit code or CISA KEV listing has been identified at time of analysis; the vendor-released fix is confirmed in libXfont2-2.0.9.

Denial-of-service in Erlang OTP's epmd daemon due to improper handling of file descriptor exhaustion allows an unauthenticated remote attacker to terminate the daemon. The vulnerability arises from calling epmd_cleanup_exit() on EMFILE/ENFILE errors instead of recovering, and can lead to permanent service failure on Debian/Ubuntu systems using systemd. No active exploitation identified (SSVC exploitation none, EPSS 0.39%) and no public exploit code is known.

In the Linux kernel, the following vulnerability has been resolved: USB: chaoskey: Fix slab-use-after-free in chaoskey_release() The chaoskey driver has a use-after-free bug in its release routine. If the user closes the device file after the USB device has been unplugged, a debugging log statement will try to access the usb_interface structure after it has been deallocated: BUG: KASAN: slab-use-after-free in dev_driver_string (drivers/base/core.c:2406) Read of size 8 at addr ffff888168e8a0b8 by task chaoskey_raw_re/10106 Hardware name: QEMU Ubuntu 24.04 PC v2 (i440FX + PIIX, arch_caps fix, 1996), BIOS 1.16.3-debian-1.16.3-2 04/01/2014 Call Trace: <TASK> dump_stack_lvl (lib/dump_stack.c:94 lib/dump_stack.c:120) print_report (mm/kasan/report.c:378 mm/kasan/report.c:482) kasan_report (mm/kasan/report.c:595) dev_driver_string (drivers/base/core.c:2406) __dynamic_dev_dbg (lib/dynamic_debug.c:906) chaoskey_release (drivers/usb/misc/chaoskey.c:323) __fput (fs/file_table.c:510) fput_close_sync (fs/file_table.c:615) __x64_sys_close (fs/open.c:1507 fs/open.c:1492 fs/open.c:1492) do_syscall_64 (arch/x86/entry/syscall_64.c:63 arch/x86/entry/syscall_64.c:94) entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:121) The driver's last reference to the interface structure is dropped in the chaoskey_free() routine, so the code must not use the interface -- even in a debugging statement -- after that routine returns. (Exception: If we know that another reference is held by someone else, such as the device core while the disconnect routine runs, there's no problem. Thanks to Johan Hovold for pointing this out.) Since the bad access is part of an unimportant debugging statement, we can fix the problem simply by removing the whole statement.

Use-after-free in the Linux kernel's Bluetooth L2CAP subsystem can be triggered by an adjacent attacker, potentially leading to arbitrary code execution or kernel memory corruption. The flaw occurs when a channel timeout callback races with connection teardown, allowing a freed connection object to be accessed. Patched kernels are available; no active exploitation is confirmed, but the vulnerability is remotely exploitable within Bluetooth range without authentication or user interaction.

Arbitrary file read in PostCSS (npm postcss <= 8.5.11) lets an attacker who controls CSS input embed a `/*# sourceMappingURL=PATH */` comment that the PreviousMap parser dereferences directly against the local filesystem with no scheme, allowlist, or traversal check. Because the read fires under default options (no `from`, `map`, or plugins), any pipeline that processes untrusted CSS is reachable; when the resulting `JSON.parse` SyntaxError is surfaced to the caller it leaks roughly the first 10 bytes of the targeted file, and the same code path yields a file-existence oracle and a DoS primitive. Publicly available exploit code exists (POC), but there is no public evidence of active exploitation, and EPSS is low at 0.50% (40th percentile).

Denial of service in Gitea's Debian package registry (all versions before 1.27.0) lets any authenticated user with package write access crash the server with a single .deb upload. The parser's ParseControlFile decompresses control.tar.gz with no size limit and tees the entire stream into a strings.Builder, so an 83 MB payload expands to 16+ GB and exhausts memory; a companion O(N²) string concatenation on Description continuation lines can independently saturate a CPU core. A complete working proof-of-concept is published in the advisory, so publicly available exploit code exists, though there is no public exploit identified as used in active attacks and the issue is not in CISA KEV.

Stack-based buffer overflow in rsyslog's optional mmpstrucdata module allows a remote, unauthenticated attacker to crash the rsyslog process and interrupt log collection, with theoretical code execution on weakly hardened builds. Affected versions span rsyslog 7.5.4 through 8.2605.x; the fix is included in 8.2606.0. No public exploit code or active exploitation has been confirmed, but the crash was independently reproduced over both TCP and UDP on Debian 12 with rsyslog 8.2302.0, and the corrective PR (#6991) was merged publicly before coordinated disclosure completed.

Poisoned-pipeline (pwn request) code execution in the Meshtastic firmware repository's GitHub Actions CI allows any external user to run arbitrary code inside the project's trusted CI context prior to firmware 2.7.21.1370b23. Because main_matrix.yml (and related workflows) triggered on pull_request_target while the check, build, and build-debian-src jobs checked out and executed code from the attacker's fork with no approval gate, a pull request from an author_association of NONE automatically ran attacker-controlled files with access to repository secrets and an elevated GITHUB_TOKEN. This is CWE-94 code injection scoring CVSS 10.0; no public exploit identified at time of analysis, though a linked Google Drive writeup may contain proof-of-concept detail.

Local denial-of-service in the Linux kernel's f2fs filesystem lets an attacker who can mount a crafted f2fs image trigger a kernel panic (BUG_ON at fs/f2fs/file.c:845) during inode truncation. A corrupted inode containing a direct node whose footer reuses the inode's own ino/nid causes ADDRS_PER_PAGE to return an inode-sized address count (923), driving ofs_in_node into the out-of-bounds addr[923..1017] range and crashing the host. Discovered via syzkaller fuzzing; no public exploit identified at time of analysis and EPSS exploitation probability is low (0.15%, 5th percentile).

OS command injection in the systeminformation Node.js library (npm, versions <= 5.31.6) lets an attacker who can influence a `source`d path in Debian/Ubuntu `interfaces(5)` files run arbitrary shell commands with the privileges of any process that calls `networkInterfaces()`. The unsanitized path token parsed from `/etc/network/interfaces` (and files it recursively sources) is interpolated into a `cat ... | grep` shell string executed via `execSync()`, so metacharacters in the path break out and execute commands. Publicly available exploit code exists (working PoC in the advisory); no public exploit is identified as being used in active exploitation. Fixed in 5.31.7.

Arbitrary file write in yutu, a Go-based YouTube CLI and MCP server, lets any caller of the built-in `caption-download` MCP tool place attacker-controlled bytes at any filesystem path the yutu process can write to. The vulnerable `Caption.Download()` calls `os.Create()` on the caller-supplied `file` parameter, bypassing the `YUTU_ROOT` (`pkg.Root`/`os.OpenRoot`) confinement that every other caption write path enforces. Publicly available exploit code exists (a self-contained Docker PoC ships with the advisory); the flaw is not in CISA KEV and no active exploitation is reported, and the vendor has released a fixed version (0.10.9-dev1).

Local privilege escalation in Linuxfabrik Monitoring Plugins allows the nagios monitoring user to obtain a root shell on Debian systems by injecting arbitrary arguments into an unrestricted apt-get sudoers rule. Deployments running version 5.0.0 or earlier with the bundled Debian.sudoers file installed are affected; exploitation requires prior compromise of the nagios account. A public proof-of-concept is documented in the vendor advisory, and a confirmed fix is available in version 5.1.0 released 2026-05-30.

Privilege escalation and project-restriction bypass in Incus before 7.2.0 lets a low-privileged user with access to an unrestricted project craft an instance whose snapshot carries malicious low-level hooks (raw.lxc/raw.qemu), then move and restore it into a project protected by restricted.containers.lowlevel=block, executing arbitrary commands on the host as root. The snapshot-restore path fails to re-enforce the lowlevel restriction, so the security control is silently bypassed. Publicly available exploit code exists (full PoC published in the vendor GHSA advisory); there is no public evidence of active exploitation.

Null pointer dereference in the Linux kernel's BPF LWT (Lightweight Tunnel) subsystem allows a local user with BPF syscall privileges to crash the kernel via a denial-of-service. The flaw exists because bpf_prog_test_run_skb constructs a socket buffer (skb) without initializing skb->_skb_refdst; when a BPF program invokes the bpf_lwt_xmit_push_encap helper through this test path, the subsequent dereference of skb_dst(skb)->dev hits a null pointer and triggers a kernel panic. No public exploit or CISA KEV listing exists, and EPSS is very low at 0.18% (7th percentile), consistent with the specialized access required.

PowerPC 64-bit (ppc64le/Book3S64) Linux kernel systems from version 5.8 onward can be crashed via a local race condition between concurrent move_pages() and munmap() operations on Transparent Huge Page (THP) PMD-backed memory. During the PMD migration sequence, a narrow window exists where the PMD entry holds a migration swap entry with pmd_present=0; a concurrent munmap() acquiring the same page table lock during this window triggers a VM_BUG_ON() assertion in pmdp_huge_get_and_clear_full(), causing a kernel panic. No public exploit is identified and EPSS stands at 0.15% (5th percentile), but the DoS impact is confirmed by upstream kernel developers with real hardware traces on IBM POWER10 and is patched in stable releases 7.0.10 and 7.1.

Local use-after-free in the Linux kernel's AF_UNIX subsystem occurs when SOCKMAP (BPF socket map) redirects skbs carrying inflight file descriptors, hiding them from the AF_UNIX garbage collector and breaking the Tarjan-based GC's assumption that unix_edge.successor stays alive — producing a slab use-after-free in unix_del_edges() plus inflight-socket leaks and incorrect fdinfo accounting. Affected are kernels supporting SOCKMAP/sk_psock redirect (roughly 5.15 through pre-fix 7.x trees); a local attacker who can set up SOCKMAP redirection and pass SCM_RIGHTS descriptors can corrupt kernel memory toward privilege escalation or denial of service. There is no public exploit identified at time of analysis, EPSS is low (0.16%, 6th percentile), and it is not in CISA KEV.

Ceph filesystem client in the Linux kernel panics with a BUG_ON assertion in __ceph_build_xattrs_blob() due to a race condition that leaves a pre-calculated xattr blob size stale, introduced by an earlier commit that restructured size computation order. All kernels from Linux 6.0 onward running a CephFS mount are affected, with fixed releases available at 6.12.91, 6.18.33, 7.0.10, and 7.1. A local low-privileged user with access to a Ceph-mounted filesystem can trigger a kernel panic causing complete system denial of service; no public exploit code or CISA KEV listing exists at time of analysis.

Local privilege escalation via command injection in Glances 4.5.5_dev1 and earlier allows users with libvirt domain-creation rights to execute arbitrary commands as the Glances process owner (typically root on hypervisor hosts). The flaw lives in the KVM/QEMU monitoring plugin, where VM domain names parsed from `virsh list --all` are interpolated into command strings handled by `secure_popen()`, which intentionally treats `&&`, `|`, and `>` as control operators. No public exploit identified at time of analysis, but a detailed PoC accompanies the GHSA-v5r2-qh84-fjx5 advisory.

Local privilege escalation / memory corruption in the Linux kernel's SO_REUSEPORT BPF handling allows a local user to trigger a use-after-free (vmalloc out-of-bounds read) by replacing a classic BPF (cBPF) reuseport program via setsockopt() while another thread routes UDP traffic through the same reuseport group. Because the cBPF program is freed immediately by sk_reuseport_prog_free() without waiting for an RCU grace period, in-flight readers in reuseport_select_sock() dereference freed memory; the fix defers freeing until after one RCU grace period. No public exploit identified at time of analysis and EPSS probability is low (0.17%), but the bug is confirmed and fixed upstream.

Authentication bypass in Bitnami Cassandra container images (4.0.x, 4.1.x, 5.0.x lines) allows remote attackers to access the database as superuser using the built-in cassandra:cassandra credentials, even when operators configure a custom administrator via CASSANDRA_USER. The container init script provisions the new superuser but, in certain scenarios, fails to drop the default account, leaving an unintended privileged login path. No public exploit identified at time of analysis, but the trivially guessable default credentials make discovery and abuse straightforward once the CQL port is reachable.

Hardcoded default credentials in Bitnami MariaDB Galera container images and Helm chart expose all default deployments to unauthenticated replication status enumeration over the network. The health-check user `monitor` with password `monitor` is granted REPLICATION CLIENT privileges from any host (`%`), meaning any network-accessible cluster running affected images (10.6.x through 12.3.x) or Helm chart prior to 18.3.0 can be accessed using these publicly known credentials. The Helm chart structurally prevented operators from overriding these credentials at deploy time, making the exposure pervasive across all standard chart-based deployments; no public exploit has been identified and the vulnerability is not in CISA KEV at time of analysis.

Unauthenticated call-control abuse in pipecat-ai development runner (>=0.0.77, <1.4.0) allows remote attackers reaching an exposed `/ws` telephony WebSocket to inject an attacker-controlled `callSid` that the server then submits to Twilio, Telnyx, or Plivo REST APIs using the operator's own credentials, forcibly terminating victim calls. Publicly available exploit code exists (a full Dockerized PoC is published in the GHSA advisory) and the maintainers shipped a fix in v1.4.0; no CISA KEV listing at time of analysis.

Server-side request forgery in Gotenberg v8.33.0 and earlier allows remote unauthenticated attackers to coerce the server into making arbitrary outbound HTTP(S) requests and to disclose local image files by uploading a crafted DOCX to the /forms/libreoffice/convert endpoint. The flaw stems from LibreOffice automatically resolving external and local resources referenced in <img> tags during conversion, exposing internal networks and on-disk image files. Publicly available exploit code exists via the vendor's GHSA-2mrg-35hw-x3x9 advisory, though no CISA KEV listing or EPSS data is available at time of analysis.

Path traversal in BBOT's unarchive internal module enables a malicious archive hosted by attacker-controlled infrastructure to write files outside the intended extraction directory when BBOT runs on systems with GNU tar < 1.34. This vulnerability is a residual gap from CVE-2025-10284, which resolved git-specific RCE vectors but left the underlying archive extraction path validation entirely unimplemented, relying instead on inconsistent external tool behavior across platforms. No public exploit has been identified at time of analysis; the CVSS vector (AC:H/UI:R) constrains real-world risk to BBOT operators actively scanning attacker-controlled targets on affected OS distributions, but the high integrity impact (I:H) and zero privilege requirement (PR:N) are significant for red-team and CI/CD BBOT deployments running on legacy base images.

Arbitrary symlink creation in Debusine's mergeuploads task allows remote unauthenticated attackers to overwrite files accessible to the worker process. Affected versions 0.12.0 through 0.14.8 fail to sanitize fully user-controlled paths embedded in Debian manifest files (.dsc and .changes), enabling path traversal via CWE-59 link-following. Despite a CVSS network-accessible, no-auth vector, real-world risk is bounded by deployment context - exploitation requires the ability to submit packages to a Debusine instance. No public exploit identified at time of analysis and EPSS sits at 0.02% (4th percentile).

Missing authorization controls on Debusine's artifact relationship endpoints allow remote actors to create and delete artifact relationships without holding the necessary write permissions, affecting all versions from 0.2.0 through 0.14.5. The only implicit gatekeeping was visibility of the referenced artifacts - not modification rights - a CWE-862 (Missing Authorization) flaw that exposes both data integrity and limited confidentiality of build artifact metadata. No public exploit code exists and EPSS places exploitation probability at 0.01% (3rd percentile), reflecting the tool's niche deployment scope within Debian build infrastructure; the vulnerability is not confirmed actively exploited (CISA KEV).

Denial of service in the Linux kernel flow dissector allows remote attackers to trigger an unaligned access exception by sending crafted PPPoE Protocol Field Compression (PFC) frames to an Ethernet interface on alignment-sensitive architectures such as MIPS. The flaw affects kernels from 6.0 onward where Receive Packet Steering (RPS) or similar features invoke the flow dissector on incoming frames, and no PPPoE session needs to be active on the targeted interface. EPSS is very low (0.02%) and there is no public exploit identified at time of analysis, but the vendor has released patched kernel versions.

Authenticated zone-file injection in Froxlor <=2.3.6 allows a customer with DNS editing enabled to inject newline characters into TXT record content via the DomainZones.add API, breaking out of the record line in the generated BIND zone file and injecting arbitrary BIND directives ($INCLUDE, $GENERATE) or DNS records (A, MX, CNAME). The flaw is an incomplete fix for CVE-2026-30932, which sanitized LOC/RP/SSHFP/TLSA records but left TXT handling reliant only on Dns::encloseTXTContent(), which strips no control characters. Publicly available exploit code exists (detailed PoC including a Python script in the GHSA advisory), but there is no public exploit identified at time of analysis in CISA KEV and no EPSS score was provided.

Out-of-bounds read in the Linux kernel's IPv6 routing subsystem (fib6_add_rt2node) allows a local user with network configuration privileges to trigger memory corruption when adding routes that use the RTA_NH_ID nexthop attribute. The flaw was discovered by syzkaller and confirmed via KASAN slab-out-of-bounds reports. No public exploit identified at time of analysis, and EPSS is very low (0.02%), but the bug is in mainline kernel code paths reachable through netlink and is fixed across multiple stable trees.

Authorization bypass in Froxlor 2.3.6 allows an authenticated low-privileged customer with shell delegation enabled to assign an arbitrary login shell (e.g., /bin/bash) to an FTP account, escaping the administrator-defined system.available_shells whitelist. On default Debian-based deployments using nssextrausers, a root-owned cron job propagates the attacker-chosen shell into /var/lib/extrausers/passwd, converting the FTP-only account into an interactive host shell account. Publicly available exploit code exists (POC published in the GHSA advisory); no public exploit identified at time of analysis as being actively used in the wild.

Authentication-context bypass in pam_usb before 0.9.0 lets a person holding an enrolled USB device authenticate over SSH while the module's deny_remote protection wrongly classifies the connection as a local terminal session. The root cause is an incomplete check of the utmpx ut_addr_v6 field that misreads IPv4-mapped IPv6 addresses (::ffff:x.x.x.x) as having no remote address, which is the normal way Debian and Ubuntu record incoming IPv4 SSH connections when sshd listens on the IPv6 wildcard. There is no public exploit identified at time of analysis and the CVE is not in CISA KEV, but the operation needed to trigger it is trivial once the operator possesses a registered token.

pam_usb prior to 0.9.0 crashes under memory pressure due to assert()-based OOM guards in src/mem.c that are silently stripped by standard distribution build flags, enabling a local denial-of-service against authentication subsystems. Any allocation failure in xmalloc(), xrealloc(), or xstrdup() returns NULL, which every caller then dereferences unconditionally - the intended abort-before-dereference guarantee exists only in debug builds, not in Debian, Fedora, or Arch Linux packages that define -DNDEBUG via CFLAGS. A local attacker who can induce memory pressure at authentication time causes the PAM module to crash, locking all users out of sudo and login for the duration of the crash. No public exploit has been identified at time of analysis and the vulnerability is not listed in the CISA KEV catalog.

Kernel NULL pointer dereference in the Linux AppArmor security module allows a local low-privileged user to crash the system by reading an apparmorfs symbolic link under a specific runtime configuration sequence. The flaw exists in rawdata_get_link_base, where profile->rawdata->name is dereferenced without first verifying that rawdata is non-NULL after a profile replacement clears it. No public exploit exists and EPSS stands at 0.02%, though the crash is fully reproducible from the conditions documented in the commit description.

Improper lock release in the Linux kernel ksmbd subsystem (in-kernel SMB server) allows a local low-privileged user to trigger a deadlock by inducing error paths in `ksmbd_vfs_kern_path_locked` where `ksmbd_vfs_kern_path_end_removing()` is never called to balance the corresponding `ksmbd_vfs_kern_path_start_removing()`. Affected kernel versions span multiple stable branches from 5.15 through 6.17. No public exploit or active exploitation is known; EPSS stands at 0.02% (7th percentile), confirming low real-world exploitation probability.

Kernel memory corruption in the Linux iWARP Connection Manager (RDMA/iwcm) subsystem can crash systems running RDMA workloads on iWARP-capable hardware such as Intel E830 adapters. The bug, introduced by commit e1168f0 ('RDMA/iwcm: Simplify cm_event_handler()'), causes workqueue list corruption when iwcm_work items from a free list are reused while still queued, triggering a kernel BUG and panic. No public exploit identified at time of analysis and EPSS exploitation probability is very low (0.02%, 5th percentile), despite a CVSS base score of 9.8.

Stack-based buffer overflow in libsolv's Debian metadata parser allows remote, unauthenticated attackers to cause a denial of service by serving maliciously crafted Debian repository metadata containing SHA384 or SHA512 checksum tags. The root cause, confirmed by the GitHub PR #616 diff, is a statically allocated 65-byte stack buffer in `ext/repo_deb.c` sized only for SHA256 digests, which is overflowed by the larger SHA384 (96 hex chars) and SHA512 (128 hex chars) values. No active exploitation has been confirmed (not in CISA KEV) and no public exploit code has been identified at time of analysis; an upstream fix is available as an open pull request.

Arbitrary file read in the esm.sh CDN build server (Go implementation, versions ≤ 137) lets a remote attacker exfiltrate sensitive host files by publishing a malicious npm package. The esbuild plugin re-maps module paths using the package.json 'browser' field without re-validating the sandbox, so '../' sequences escape the package directory and file contents are returned inside the bundled JS and source-map sourcesContent. A working PoC exists (published npm package chess-sec-utils1) reading /etc/hostname, /etc/os-release and /etc/environment, but there is no public exploit identified as actively used and EPSS risk is low (0.04%).

Path traversal in Open WebUI's file upload mechanism allows authenticated attackers to write and subsequently delete arbitrary files on the server filesystem. Discovered by Taylor Pennington of KoreLogic, this vulnerability affects the /ollama/models/upload API endpoint where unsanitized filename parameters enable directory traversal using dot-segments. The vulnerability requires low-privilege authentication (PR:L) and has straightforward exploitation (AC:L), confirmed by a published GitHub security advisory (GHSA-j3fw-wc48-29g3) with working proof-of-concept code. Vendor-released patch available in version 0.6.10. No evidence of active exploitation (not in CISA KEV) at time of analysis.

Arbitrary file write in Open WebUI (formerly Ollama WebUI) 0.1.105 through 0.1.123 lets a user of the chat interface abuse the /rag/api/v1/doc document-upload endpoint to plant attacker-controlled files anywhere the web server process can write. Because the uploaded filename is used verbatim to build the destination path, dot-segment traversal (../../) escapes the uploads directory, enabling code execution by dropping a malicious pickled model or an SSH authorized_keys file. Publicly available exploit code exists (a working cURL PoC is published in the KoreLogic/GHSA advisory), but there is no confirmed active exploitation; EPSS is low at 0.06% (17th percentile).

Improper authorization in Open WebUI (formerly Ollama WebUI) versions up to and including 0.1.123 lets self-registered accounts stuck in the 'pending' role make authenticated API calls that should be restricted to approved 'user'/'admin' accounts. Because the API only checks that a JWT is valid (get_current_user) and never verifies the role, an attacker who signs up when open registration is enabled can query connected LLMs and enumerate models despite an admin never approving them. This is a known-PoC issue with publicly available exploit code from the KoreLogic disclosure; EPSS is low (0.06%, 20th percentile) and it is not in CISA KEV.

Integer overflow in Linux kernel's libceph authentication handler enables remote memory corruption and potential system crash against unpatched systems. A malicious Ceph monitor can send a specially crafted CEPH_MSG_AUTH_REPLY message with payload_len exceeding INT_MAX, causing ceph_handle_auth_reply() to underflow a pointer and trigger out-of-bounds memory access. This allows remote unauthenticated attackers to potentially read sensitive kernel memory (high confidentiality impact) or crash the kernel (high availability impact) on systems using Ceph storage. CVSS 9.1 (Critical) reflects network attack vector with no authentication or user interaction required. EPSS score of 0.02% (7th percentile) suggests low observed exploitation likelihood. Vendor patches available for all affected kernel series (5.10.253, 5.15.203, 6.1.167, 6.6.130, 6.12.78, 6.18.19, 6.19.9, 7.0), but no active exploitation confirmed via CISA KEV.

Denial of service via kernel warning in MPTCP path manager occurs when combining endpoint removal with fullmesh and flag-setting operations through netlink in the Linux kernel. A local attacker with low privileges can trigger a WARNING in net/mptcp/pm_kernel.c:1074 by sending a crafted sequence of netlink commands, causing the system to emit a kernel warning and potentially become unstable. No known public exploit code exists, but the low CVSS (5.5) and minimal EPSS (0.03%) indicate this is a local DoS with limited real-world impact.

Integer overflow in the Linux kernel's l2tp subsystem allows a local low-privileged attacker to cause a denial of service by sending oversized packets through a PPPoL2TP socket with UDP encapsulation. In l2tp_xmit_core() (net/l2tp/l2tp_core.c:1293), the UDP length field - constrained to 16 bits - is assigned a packet length value without any bounds check, silently truncating values exceeding 65535 bytes and producing malformed UDP frames on the wire or triggering a kernel WARN. No public exploit has been identified at time of analysis and the vulnerability is not listed in CISA KEV, but a syzbot reproducer is publicly documented in the upstream patch discussion.

Blind server-side request forgery in Incus allows authenticated users to trigger arbitrary HEAD requests to internal or external endpoints during image import preflight validation, bypassing the restricted.images.servers project restriction. While the actual image download is blocked by project policies, the preflight HEAD request executes before validation occurs, enabling attackers to probe internal services, cloud metadata endpoints, or unroutable address space reachable from the Incus host. No public exploit code identified at time of analysis, though proof-of-concept reproduction is documented in the advisory.

Kernel denial of service via crafted btrfs metadata allowing local attackers to trigger an unguarded BUG_ON() condition during relocation recovery at mount time. The vulnerability arises when a root item on disk contains a non-zero drop_progress with zero drop_level, an invalid state that should not exist but lacks validation on read. CVSS 5.5 reflects local attack vector and availability impact; EPSS 0.02% indicates minimal real-world exploitation likelihood.

Memory leak in the Linux kernel's __mmap_region() function allows a local low-privileged user to incrementally exhaust kernel slab memory, potentially causing denial of service. Introduced as a regression by commit 605f6586ecf7, the bug leaves a shmem file object unreferenced when mmap() on /dev/zero with MAP_SHARED fails mid-way through VMA setup - the replacement file allocated by shmem_zero_setup_desc() is never released via fput(). No public exploit exists and EPSS sits at 0.02%, making this a low-urgency but genuine availability risk for affected kernel versions; patched upstream in stable commits 61fc8eaf and 894f99eb.

SQL injection in NocoBase's @nocobase/database package allows authenticated users with record-creation privileges to execute arbitrary SQL queries and extract database credentials. The vulnerability exists in the queryParentSQL() function, which constructs recursive Common Table Expression (CTE) queries using string concatenation instead of parameterized queries when processing tree collections with string primary keys. An attacker can inject malicious SQL by creating records with crafted primary key values, triggering the vulnerability when recursive eager loading occurs. Successful exploitation leads to full database compromise, with confirmed extraction of administrator credentials (emails and password hashes) in testing against PostgreSQL. On databases where the service account has elevated privileges, attackers can achieve operating system command execution via PostgreSQL's COPY...TO PROGRAM feature. Vendor patch available via GitHub PR #9133.

Path traversal vulnerability in Poetry's tar extraction function allows arbitrary file writes when processing untrusted source distributions on Python 3.10.0-3.10.12 and 3.11.0-3.11.4, where the tarfile.data_filter safety mechanism is absent or broken. The vulnerability is triggered during dependency resolution (poetry add --lock) or installation before the build backend executes, enabling attackers to write files outside the intended extraction directory via crafted tar member paths, symlinks, or hardlinks in malicious sdists.

Out-of-bounds memory write in Linux kernel iavf (Intel Adaptive Virtual Function) driver allows local authenticated attackers with low privileges to achieve high confidentiality, integrity, and availability impact via race condition during concurrent ethtool operations. The vulnerability stems from inconsistent use of queue counters (real_num_tx_queues vs num_active_queues vs num_tx_queues) across ethtool statistics functions, enabling memory corruption when changing network channels via 'ethtool -L' while simultaneously querying statistics with 'ethtool -S'. Patches available for kernel versions 6.12.80, 6.18.21, 6.19.11, and 7.0. EPSS exploitation probability is low (0.02%, 5th percentile) with no public exploit or active exploitation identified at time of analysis.

Use-after-free vulnerability in the Linux kernel's Bluetooth HIDP subsystem allows local attackers to trigger a kernel crash or potentially execute arbitrary code by failing to properly release L2CAP connection references when user callbacks are invoked. The flaw affects all Linux kernel versions in the CPE range and has been resolved through reference counting fixes in the L2CAP connection cleanup path; no public exploit code is currently identified, but the vulnerability requires local access to trigger via Bluetooth device manipulation.

Use-after-free in Linux kernel ksmbd SMB server allows local or remote attackers to read freed memory and potentially achieve denial of service or code execution via compound SMB2 requests that reuse a tree connection after it has been disconnected and its associated share_conf structure freed. The vulnerability exists because smb2_get_ksmbd_tcon() bypasses state validation checks when reusing connections in compound requests, enabling subsequent commands to dereference already-freed share_conf pointers. No CVE severity metrics are available, but KASAN confirms memory corruption is triggered in smb2_write operations during tree disconnect sequences.

Use-after-free in Linux kernel's ksmbd SMB server allows remote attackers to crash the kernel or potentially execute code via malicious SMB2 DURABLE_REQ_V2 replay operations. The vulnerability occurs when parse_durable_handle_context() unconditionally reassigns file handle connection pointers during replay operations, causing stale pointer dereferences when the reassigned connection is subsequently freed. A KASAN report confirms the use-after-free in spin_lock operations during file descriptor closure, triggered during SMB2 connection handling in the ksmbd-io workqueue. No public exploit code or active exploitation has been confirmed at time of analysis.

SQL injection in NocoBase plugin-workflow-sql through version 2.0.8 allows authenticated workflow users to execute arbitrary database queries. The vulnerable SQLInstruction class performs unparameterized string substitution of template variables (e.g., {{$context.data.fieldName}}) directly into raw SQL statements, enabling attackers to break out of string literals and inject malicious SQL commands. Publicly available exploit code exists demonstrating UNION-based injection to extract database credentials and system information. With default Docker deployments granting superuser database privileges, attackers gain full read/write access to the database including credential extraction, data modification, and table deletion capabilities.

Out-of-bounds read and write in Linux kernel AppArmor DFA verification allows local authenticated users to cause memory corruption and potential privilege escalation. Affects Linux kernel 4.17+ through 6.12.x, actively patched in stable branches 6.6.130, 6.12.77, 6.18.18, and 6.19.8. The vulnerability occurs during AppArmor policy loading when malformed DFA structures bypass bounds checking in verify_dfa(). EPSS score of 0.02% suggests low exploitation probability in the wild; no public exploit code or CISA KEV listing identified. Ubuntu rates this medium priority with patches released via USN-8152-1.

Out-of-bounds memory read in Linux kernel AppArmor DFA matching affects kernel versions since 4.17 due to macro side-effect bug. The match_char() macro evaluates its character parameter multiple times when called with *str++, causing the pointer to advance past buffer boundaries during differential encoding chain traversal. Ubuntu has released patches (USN-8152-1) with fixes available in kernels 6.6.130, 6.12.77, 6.18.18, 6.19.8, and 7.0-rc4. EPSS score of 0.02% (5th percentile) indicates very low exploitation probability, and no public exploit or CISA KEV listing exists. CVSS 7.8 reflects local high-privilege impact, but real-world risk is limited to authenticated local users who can trigger AppArmor path permission checks.

Out-of-bounds read in WebCodecs component of Google Chrome prior to version 146.0.7680.178 allows remote attackers to read arbitrary memory contents via specially crafted HTML pages. The vulnerability affects all Chrome versions below the patched release and requires only HTML delivery (no authentication); exploitation could disclose sensitive data from the browser process memory, though the Chromium project assessed this as Medium severity.

Information disclosure in Google Chrome's WebGL implementation prior to version 146.0.7680.178 allows remote attackers to extract potentially sensitive data from process memory by serving a crafted HTML page. The vulnerability affects all Chrome versions before the patched release and requires only user interaction (visiting a malicious webpage) to trigger memory disclosure via WebGL rendering.

Use-after-free in Chrome's compositing engine allows remote attackers who have compromised the renderer process to escape the sandbox via crafted HTML pages in Google Chrome prior to version 146.0.7680.178. This high-severity vulnerability requires prior renderer compromise but enables privilege escalation from the sandboxed renderer to system-level access, making it a critical sandbox bypass vector. Vendor-released patch addresses the issue in Chrome 146.0.7680.178 and later.

Use-after-free in Google Chrome's Navigation component prior to version 146.0.7680.178 enables sandbox escape for attackers who have already compromised the renderer process, allowing them to potentially execute arbitrary code with elevated privileges via a malicious HTML page. Chromium rates this as high severity; patch availability confirmed from vendor.

Use-after-free in Chrome's WebView on Android prior to version 146.0.7680.178 allows a remote attacker with a compromised renderer process to escape the sandbox via crafted HTML, potentially leading to arbitrary code execution outside the browser's security boundary. This vulnerability requires prior renderer compromise but eliminates a critical containment layer, classified as High severity by Chromium.

Remote code execution in Google Chrome prior to version 146.0.7680.178 via use-after-free vulnerability in the Dawn graphics library allows unauthenticated remote attackers to execute arbitrary code through a crafted HTML page. The vulnerability affects all Chrome versions below the patched release and carries high severity per Chromium's assessment.

Remote code execution in Google Chrome prior to 146.0.7680.178 via use-after-free vulnerability in Dawn graphics subsystem allows an attacker who has already compromised the renderer process to execute arbitrary code through a crafted HTML page. This vulnerability requires prior renderer compromise but presents significant risk in multi-process exploitation chains; vendor has released patched version 146.0.7680.178 to address the issue.

Information disclosure in ANGLE (graphics abstraction layer) within Google Chrome prior to version 146.0.7680.178 enables remote attackers to leak cross-origin data through crafted HTML pages. The vulnerability affects all Chrome versions before the patched release and requires only network access and user interaction (visiting a malicious page), posing a moderate real-world risk to users who may inadvertently access attacker-controlled content.

Out-of-bounds read in WebCodecs functionality in Google Chrome prior to version 146.0.7680.178 allows remote attackers to read arbitrary memory contents via a crafted HTML page. The vulnerability affects all Chrome versions before the patched release and requires only user interaction (visiting a malicious webpage) to trigger. No public exploit code or active exploitation has been confirmed at time of analysis.

Remote code execution in Google Chrome on Android via use-after-free vulnerability in Web MIDI allows unauthenticated remote attackers to execute arbitrary code through a crafted HTML page. The vulnerability affects Chrome versions prior to 146.0.7680.178 and carries high severity per Chromium's security classification. A vendor-released patch is available.

Integer overflow in ANGLE (Google's OpenGL abstraction layer) in Chrome on Windows before version 146.0.7680.178 enables out-of-bounds memory writes if the renderer process is compromised, allowing an attacker to execute arbitrary code with renderer privileges. The vulnerability requires prior renderer process compromise, limiting the immediate attack surface but representing a critical post-compromise escalation vector. Chromium severity is rated High; patch availability confirms vendor remediation.

Information disclosure in Google Chrome's WebUSB implementation prior to version 146.0.7680.178 allows remote attackers to extract sensitive data from process memory by delivering a crafted HTML page, exploiting insufficient policy enforcement in the WebUSB API. The vulnerability affects all Chrome versions before 146.0.7680.178 across all platforms. No public exploit code or active exploitation has been confirmed at the time of this analysis.

Remote code execution in ANGLE (Almost Native Graphics Layer Engine) within Google Chrome on macOS prior to version 146.0.7680.178 allows unauthenticated remote attackers to execute arbitrary code by crafting a malicious HTML page that triggers a heap buffer overflow. This vulnerability affects all Chrome versions below the patched release and poses an immediate risk to macOS users who visit compromised or malicious websites.

Integer overflow in Google Chrome's Codecs component prior to version 146.0.7680.178 enables remote code execution and arbitrary memory read/write operations when a user visits a malicious HTML page. The vulnerability affects all versions before the patch release and requires no user interaction beyond visiting a crafted webpage. Chromium security team classified this as High severity; no public exploit code or active exploitation has been confirmed at the time of analysis.

Remote code execution in NocoBase Workflow Script Node (npm @nocobase/plugin-workflow-javascript) allows authenticated low-privilege attackers to escape Node.js vm sandbox and execute arbitrary commands as root inside Docker containers. The vulnerability exploits exposed WritableWorkerStdio stream objects in the sandbox console to traverse the prototype chain, access the host-realm Function constructor, load unrestricted Node.js modules (child_process), and spawn system commands. Confirmed exploited with reverse shell access, database credential theft (DB_PASSWORD, INIT_ROOT_PASSWORD), and arbitrary filesystem operations. EPSS data not available; public exploit code exists with detailed proof-of-concept demonstrating root shell access in nocobase/nocobase:latest Docker image. Critical 10.0 CVSS score reflects network-exploitable, low-complexity attack with complete confidentiality, integrity, and availability impact plus scope change (container escape implications).

Mattermost versions 11.2.x through 11.2.2, 10.11.x through 10.11.10, 11.4.0, and 11.3.x through 11.3.1 fail to properly restrict team-level access during remote cluster membership synchronization, allowing a malicious remote cluster to grant users access to entire private teams rather than limiting access to only shared channels. An authenticated attacker controlling a federated remote cluster can send crafted membership sync messages to trigger unintended team membership assignment, resulting in unauthorized access to private team resources. The EPSS score of 0.03% (percentile 7%) indicates low real-world exploitation probability, and no public exploit code has been identified at time of analysis.

A validation bypass vulnerability exists in the Linux kernel's netfilter nft_set_rbtree module that fails to properly validate overlapping open intervals in packet filtering rule sets. This affects all Linux distributions running vulnerable kernel versions, allowing local or remote attackers with network configuration privileges to bypass firewall rules through malformed interval specifications. The vulnerability is classified as an information disclosure issue and has been patched upstream, though no active exploitation in the wild has been documented.

A logic error in the Linux kernel's MPTCP (MultiPath TCP) path management subsystem fails to properly track endpoint usage state when an endpoint is configured with both 'signal' and 'subflow' flags and subsequently removed. This causes a kernel warning and potential state inconsistency in the MPTCP connection management code. The vulnerability affects Linux kernel versions and is triggered through netlink socket manipulation by unprivileged users, potentially leading to denial of service or unexpected kernel behavior.

A lifecycle management vulnerability in the Linux kernel's USB NCM (Network Control Model) gadget function causes the network device to outlive its parent gadget device, resulting in NULL pointer dereferences and dangling sysfs symlinks when the USB gadget is disconnected. This affects all Linux kernel versions with the vulnerable USB gadget NCM implementation, and an attacker with local access to trigger USB gadget bind/unbind cycles can cause a kernel panic (denial of service). No CVSS vector, EPSS score, or active KEV status is available, but patches are confirmed available in the Linux stable tree.

A NULL pointer dereference vulnerability exists in tmate versions prior to 2.4.0, allowing unauthenticated remote attackers to cause a denial of service condition by crashing the application. The vulnerability has a CVSS score of 5.3 (medium severity) with low attack complexity and no privilege requirements, making it readily exploitable over the network. A patch is available from the vendor, and this issue does not compromise confidentiality or integrity-only availability.

Remote code execution in Google Chrome's Federated Credential Management (FedCM) prior to version 146.0.7680.165 enables unauthenticated attackers to execute arbitrary code within the browser sandbox through a malicious HTML page. This use-after-free vulnerability in memory management affects Chrome on all supported platforms and requires only user interaction to trigger. A patch is available in Chrome 146.0.7680.165 and later.

Out-of-bounds memory write in Google Chrome's font handling prior to version 146.0.7680.165 enables remote code execution when users visit malicious HTML pages. An unauthenticated attacker can exploit an integer overflow vulnerability to achieve complete system compromise with high integrity and confidentiality impact. Patches are available for Chrome and affected Debian systems.

Sandboxed code execution in Google Chrome's WebGPU implementation (prior to 146.0.7680.165) stems from a use-after-free memory vulnerability that can be triggered via malicious HTML pages. An unauthenticated remote attacker can exploit this to execute arbitrary code within the Chrome sandbox without user interaction beyond viewing a crafted webpage. A patch is available for affected users.

This vulnerability is an out-of-bounds memory read flaw in the WebAudio API implementation within Google Chrome prior to version 146.0.7680.165. A remote attacker can craft a malicious HTML page to trigger the vulnerability and read sensitive memory contents, leading to information disclosure. Although no CVSS score or EPSS data is provided, the Chromium security severity is rated as High, and the vulnerability affects all users of vulnerable Chrome versions until patching.

Sandbox escape in Google Chrome prior to version 146.0.7680.165 via a use-after-free vulnerability in the Dawn graphics component enables remote attackers to execute arbitrary code when users visit malicious HTML pages. The vulnerability affects multiple platforms including Debian systems and requires only user interaction to trigger, bypassing Chrome's sandbox isolation. A patch is available to remediate this high-severity memory corruption flaw.

Google Chrome's WebGL implementation contains a heap buffer overflow that enables remote attackers to read arbitrary memory by serving a specially crafted HTML page to users prior to version 146.0.7680.165. This network-based vulnerability requires only user interaction and affects Chrome on all platforms, granting attackers access to sensitive data in the browser's memory. A patch is available and should be applied immediately given the high severity and potential for exploitation.

Out of bounds memory read in Google Chrome's CSS parser prior to version 146.0.7680.165 allows remote attackers to access sensitive memory contents through a malicious HTML page. The vulnerability requires user interaction and affects Chrome on multiple platforms including Debian systems, enabling attackers to potentially leak confidential data with high impact on confidentiality and integrity.