Nomad's exec2 task driver vulnerable to arbitrary file read/write on client host through symlink attack

HashiCorp Nomad’s exec2 task driver prior to 0.1.2 is vulnerable to arbitrary file read and write on the client host as the Nomad process user through a symlink attack. This vulnerability (CVE-2026-8052) is fixed in version 0.1.2 of the exec2 task driver.

GPT-Pilot thru commit 0819827ce20346ef5f25b3fe29293cb448840565 (2025-09-03) contains a command injection vulnerability (CWE-78) in the Executor.run() method. During project execution, when the system prompts the user to confirm or modify a command to be run, it accepts free-text input without proper validation. The user-supplied input is directly passed to asyncio.create_subprocess_shell() for execution. This allows an attacker to replace the intended command with arbitrary shell commands, leading to remote code execution with the privileges of the GPT-Pilot process.

Insecure Deserialization in CosyVoice gRPC Server Enables Remote Code Execution

CosyVoice thru commit 6e01309e01bc93bbeb83bdd996b1182a81aaf11e (2025-30-21) contains an insecure deserialization vulnerability (CWE-502) in its gRPC server component. When the server starts, it loads the speech synthesis model from a user-specified directory using torch.load() without enabling the weights_only=True security parameter. This allows the deserialization of arbitrary Python objects via the pickle module. An attacker can exploit this by providing malicious model files within a directory. When a victim starts the gRPC server pointing to this directory, arbitrary code is executed on the victim's system during server initialization.

Insecure Deserialization in CosyVoice Model Loading Allows Arbitrary Code Execution

CosyVoice thru commit 6e01309e01bc93bbeb83bdd996b1182a81aaf11e (2025-30-21) contains an insecure deserialization vulnerability (CWE-502) in its model loading component. The framework uses torch.load() to load model weight files (e.g., llm.pt, flow.pt, hift.pt) without enabling the security-restrictive weights_only=True parameter. This allows the deserialization of arbitrary Python objects via the pickle module. An attacker can exploit this by providing a malicious model directory containing specially crafted model files. When a victim starts the CosyVoice Web UI pointing to this directory, arbitrary code is executed on the victim's system during the model loading process.

flash‑attention Training Script Allows Arbitrary Code Execution via eval Resolver

The flash-attention project thru commit e724e2588cbe754beb97cf7c011b5e7e34119e62 (2025-13-04) contains a code injection vulnerability (CWE-94) in its training script. The script registers the Python eval() function as a Hydra configuration resolver under the name eval. This allows configuration files to execute arbitrary Python code via the ${eval:...} syntax. An attacker can exploit this by providing a malicious configuration file, leading to arbitrary code execution when the training script is run with that configuration.

Fides: Privacy Request Identity Verification Bypass Vulnerability via Duplicate Detection

Fides is an open-source privacy engineering platform. From 2.75.0 to before 2.83.2, Fides deployments that enable both subject identity verification and duplicate privacy request detection are affected by a vulnerability in which an administrator can approve a privacy request whose identity was never verified. For erasure policies, this can result in unauthorized deletion of a data subject's records across every integration configured in the affected deployment. This vulnerability is fixed in 2.83.2.

Cleanuparr: Reflective CORS combined with trusted-network auth allows cross-origin admin API reads

Cleanuparr is a tool for automating the cleanup of unwanted or blocked files in Sonarr, Radarr, and supported download clients like qBittorrent. Prior to 2.9.10, Cleanuparr's global CORS policy reflects every request Origin and combines it with AllowCredentials(). When DisableAuthForLocalAddresses is enabled, the API also authenticates requests purely by source IP via TrustedNetworkAuthenticationHandler. The combination lets any website that an admin (or any user on a trusted IP) visits read authenticated API responses cross-origin — including the admin's permanent API key. This vulnerability is fixed in 2.9.10.

App may access sensitive user data due to inconsistent UI state in iOS and iPadOS

An inconsistent user interface issue was addressed with improved state management. This issue is fixed in iOS 18.7.3 and iPadOS 18.7.3, iOS 26.2 and iPadOS 26.2. An app may be able to access sensitive user data.

media: ipu6: Fix RPM reference leak in probe error paths

In the Linux kernel, the following vulnerability has been resolved: media: ipu6: Fix RPM reference leak in probe error paths Several error paths in ipu6_pci_probe() were jumping directly to out_ipu6_bus_del_devices without releasing the runtime PM reference. Add pm_runtime_put_sync() before cleaning up other resources.

A race condition was addressed with additional validation. This issue is fixed in macOS Tahoe 26.4. An app may be able to access sensitive user data.

CosyVoice thru commit 6e01309e01bc93bbeb83bdd996b1182a81aaf11e (2025-30-21) contains an insecure deserialization vulnerability (CWE-502) in its make_parquet_list.py data processing tool. The script loads PyTorch .pt files (utterance embeddings, speaker embeddings, speech tokens) using torch.load() without enabling the weights_only=True security parameter. This allows the deserialization of arbitrary Python objects via the pickle module. An attacker can exploit this by providing malicious .pt files within a data directory. When a victim processes this directory using the tool, arbitrary code is executed on the victim's system.

The flash-attention training framework thru commit e724e2588cbe754beb97cf7c011b5e7e34119e62 (2025-13-04) contains an insecure deserialization vulnerability (CWE-502) in its checkpoint loading mechanism. The load_checkpoint() function in checkpoint.py and the checkpoint loading code in eval.py use torch.load() without enabling the security-restrictive weights_only=True parameter. This allows the deserialization of arbitrary Python objects via the pickle module. An attacker can exploit this by providing a maliciously crafted checkpoint file. When a victim loads this checkpoint during model warmstarting or evaluation, arbitrary code is executed on the victim's system.

SQL Injection in MuuCMF T6 v1.9.4.20260115 allows an unauthenticated attacker to compromise the entire database, achieve unauthorized administrative access, and potentially gain remote code execution by writing malicious files to the server's file system via the keyword parameter in the /index/controller/Search.php endpoint.

XML Entity Expansion (XXE) Exploitation Causing DoS in Docling METS GBS Backend

Docling's METS GBS backend is vulnerable to XML Entity Expansion (XXE) attacks thru 2.61.0. The backend extracts and validates XML files from .tar.gz archives using etree.fromstring() without disabling entity resolution. An attacker can craft a malicious XML file with nested entity definitions (XML Bomb) and package it into a .tar.gz archive. When processed by Docling, the exponential expansion of entities during XML parsing leads to excessive resource consumption, resulting in a denial of service (DoS) condition on the system running the Docling parser.

CosyVoice thru commit 6e01309e01bc93bbeb83bdd996b1182a81aaf11e (2025-30-21) contains an insecure deserialization vulnerability (CWE-502) in its average_model.py model averaging tool. The script loads PyTorch checkpoint files (epoch_*.pt) for model averaging using torch.load() without enabling the weights_only=True security parameter. This allows the deserialization of arbitrary Python objects via the pickle module. An attacker can exploit this by providing malicious checkpoint files within a directory. When a victim uses the tool to average models from this directory, arbitrary code is executed on the victim's system.

Mail Logic Flaw Allows Remote Images During Replies in Lockdown Mode

A logic issue was addressed with improved checks. This issue is fixed in iOS 18.7.9 and iPadOS 18.7.9, macOS Sequoia 15.7.7, macOS Sonoma 14.8.7, macOS Tahoe 26.5. Replying to an email could display remote images in Mail in Lockdown Mode.

LobeHub: Cross-Site Scripting(XSS) escalate to Remote Code Execution(RCE)

LobeHub is a work-and-lifestyle space to find, build, and collaborate with agent teammates that grow with you. Prior to 2.1.48, when LobeChat processes custom tags in the Render process of src/features/Portal/Artifacts/Body/Renderer/index.tsx, if no type match is found, it will choose to call the default method, HTMLRenderer, for HTML rendering. If an attacker can induce the LLM to output content containing malicious tags, an XSS vulnerability can be created on the client side. Additionally, Lobechat's Electron main process exposes an IPC interface called runCommand, used to invoke system commands. This interface allows arbitrary command execution and does not filter the command parameter. Therefore, if an attacker can obtain a handle to window.parent.electronAPI via XSS and call the runCommand method of the IPC, the ipcMain process can execute arbitrary system commands with the current user's privileges. This vulnerability is fixed in 2.1.48.

docuFORM Managed Print Service Client 11.11c is vulnerable to a directory traversal allowing attackers to read arbitrary files via crafted url.

hfsplus: ensure sb->s_fs_info is always cleaned up

In the Linux kernel, the following vulnerability has been resolved: hfsplus: ensure sb->s_fs_info is always cleaned up When hfsplus was converted to the new mount api a bug was introduced by changing the allocation pattern of sb->s_fs_info. If setup_bdev_super() fails after a new superblock has been allocated by sget_fc(), but before hfsplus_fill_super() takes ownership of the filesystem-specific s_fs_info data it was leaked. Fix this by freeing sb->s_fs_info in hfsplus_kill_super().

io_uring/zcrx: fix user_ref race between scrub and refill paths

In the Linux kernel, the following vulnerability has been resolved: io_uring/zcrx: fix user_ref race between scrub and refill paths The io_zcrx_put_niov_uref() function uses a non-atomic check-then-decrement pattern (atomic_read followed by separate atomic_dec) to manipulate user_refs. This is serialized against other callers by rq_lock, but io_zcrx_scrub() modifies the same counter with atomic_xchg() WITHOUT holding rq_lock. On SMP systems, the following race exists: CPU0 (refill, holds rq_lock) CPU1 (scrub, no rq_lock) put_niov_uref: atomic_read(uref) - 1 // window opens atomic_xchg(uref, 0) - 1 return_niov_freelist(niov) [PUSH #1] // window closes atomic_dec(uref) - wraps to -1 returns true return_niov(niov) return_niov_freelist(niov) [PUSH #2: DOUBLE-FREE] The same niov is pushed to the freelist twice, causing free_count to exceed nr_iovs. Subsequent freelist pushes then perform an out-of-bounds write (a u32 value) past the kvmalloc'd freelist array into the adjacent slab object. Fix this by replacing the non-atomic read-then-dec in io_zcrx_put_niov_uref() with an atomic_try_cmpxchg loop that atomically tests and decrements user_refs. This makes the operation safe against concurrent atomic_xchg from scrub without requiring scrub to acquire rq_lock. [pavel: removed a warning and a comment]