Fluent Bit is a fast and lightweight logs, metrics, and traces processor for Linux, BSD, macOS, and Windows. From 0.11.0 until 5.0.8, plugins/out_forward/forward.c secure_forward_pong copies the server-controlled PONG[2] reason into the 32-byte stack buffer msg with memcpy without checking its MessagePack type or length. An attacker who controls or can impersonate an out_forward Secure Forward destination configured with Shared_Key or Empty_Shared_Key can send an oversized reason during the first handshake and overwrite stack control data. Protected builds reliably terminate, while builds without a stack canary or with a disclosure can allow remote code execution as the Fluent Bit process user. When the opt-in --supervisor mode is used, fork-only respawns preserve the canary and address layout, allowing repeated crash-or-survive probes to support code execution on a hardened build; ordinary exec-based or service-manager restarts do not preserve that state. This issue is fixed in version 5.0.8.

MISP's galaxy matrix statistics view (app/View/Users/statistics_galaxymatrix.ctp) renders the galaxy name directly into HTML output via sprintf() without any HTML encoding. An authenticated user holding the perm_galaxy_editor permission can create or modify a galaxy whose name contains arbitrary HTML or JavaScript markup. Because the value is interpolated verbatim into the page, any user who subsequently opens the galaxy matrix statistics page will have the embedded script executed in their browser context. This enables session hijacking, credential theft, data exfiltration, or the performance of arbitrary actions on behalf of the victim within the MISP application. Version affected: <2.5.47

MISP contains a stored cross-site scripting (XSS) vulnerability in the default theme's Galaxies index page. When a MISP instance detects unknown custom or default galaxy clusters during synchronization, it renders sample tag names in an informational notice directed at site administrators. In the default theme, these sample tag names were inserted into the HTML output without HTML-entity encoding. A user with tag-editor privileges could craft a malicious misp-galaxy tag containing arbitrary script. When a site administrator subsequently viewed the Galaxies index page, the crafted tag content was interpreted as executable markup in the administrator's browser session, enabling the attacker to read session data, perform actions on behalf of the administrator, or otherwise compromise the administrator's browser context. The Overmind theme already applied HTML escaping to the same values and was not affected. The vulnerability requires the attacker to possess tag-editor-level access to create the crafted tag and requires the victim (a site administrator) to view the Galaxies index page in the default theme. The impact is limited to the administrator's browser session and does not directly compromise the MISP server process. Version affected: <2.5.47

A vulnerability was determined in ST Engineering iDirect Evolution and Velocity WebServer Evolution up to 20260717. This vulnerability affects the function authorize of the file /usr/sbin/webserver of the component HTTP Header Handler. Executing a manipulation of the argument Success can lead to open redirect. It is possible to launch the attack remotely. The exploit has been publicly disclosed and may be utilized. The vendor was contacted early about this disclosure but did not respond in any way.

The Intel SEDI IPM (inter-processor mailbox) driver in drivers/ipm/ipm_sedi.c handles an inbound message interrupt in ipm_event_dispose(). It read the peer-written doorbell register, extracted the payload length with IPC_HEADER_GET_LENGTH(), and passed that length straight to sedi_ipc_read_msg() to copy the message into struct ipm_sedi_context.incoming_data_buf, without checking it against the buffer size. The doorbell length field is 10 bits wide (IPC_HEADER_LENGTH_MASK is 0x03FF), so it can encode up to 1023 bytes, while incoming_data_buf is IPC_DATA_LEN_MAX (128) bytes. The bounds check in the underlying HAL sedi_ipc_read_msg() is a DBG_CHECK that compiles away unless CONFIG_DEBUG is set, so no check remained in a production image. The doorbell register is written by the peer processor on the other side of the IPC link — for the intel_ish_5_* targets, the host CPU's ISH driver, reached through the device's memory-mapped register window. Host-side software with driver-level or raw BAR access can therefore set a length of up to 1023 and cause the interrupt handler to copy far past the destination buffer. The affected path requires an application to have registered an IPM receive callback via ipm_register_callback(), which is the driver's normal mode of use. The result is an out-of-bounds write of up to 895 bytes into static (.bss) memory, performed in interrupt context. The overflow first clobbers the rest of struct ipm_sedi_context — including the k_sem and k_mutex used by the transmit path, whose wait queues contain self-referential list pointers — and then adjacent static data, giving a kernel data-structure corruption and crash primitive. The overflowing bytes are read from registers following the message window, a portion of which are themselves peer-programmable. The fix rejects any doorbell whose encoded length exceeds IPC_DATA_LEN_MAX, logging it and acknowledging the doorbell so the peer is not left waiting.

Tautulli is a Python based monitoring and tracking tool for Plex Media Server. Prior to 2.17.2, the /search endpoint inserts its user-controlled query parameter into a JavaScript string in data/interfaces/default/search.html using manual escaping that handles quotes and slashes but not backslashes. A backslash-quote sequence can terminate the string, so an unauthenticated attacker can send a crafted link that executes script in the Tautulli web context when an authenticated user follows it. This issue is fixed in version 2.17.2.

Joplin is an open source note-taking and to-do application that organises notes and lists into notebooks. Prior to 3.6.15 and 3.7.2, Joplin accepts synchronized resource metadata whose id or file_extension contains parent-directory or path-separator characters. BaseItem.unserialize() stores the unvalidated fields, resourceFilename() concatenates them into a destination path, and ResourceFetcher writes the attacker-controlled resource blob outside the resource directory during background synchronization. An attacker with write access to a configured sync target or shared notebook can create or overwrite files at an attacker-chosen existing path without user interaction. This issue is fixed in versions 3.6.15 and 3.7.2.

MISP has a security issue that could let an attacker change threat-intelligence data through a logged-in user’s browser without that user knowingly approving the change. The affected function did not properly enforce MISP’s usual protection against forged requests. Because of this, an attacker could create a malicious webpage that silently sends a request to MISP when visited by an authenticated user. If successful, the attacker could change details of an attribute, such as its value, type, category, comment, distribution settings, or related timestamps. The attack requires the victim to already be logged in to MISP and to visit an attacker-controlled page. The main impact is unauthorized modification of threat-intelligence data, which could lead to incorrect indicators, wrong classifications, or altered sharing settings and reduce confidence in the accuracy of the information stored in MISP. Version affected: <2.5.47

nvm resolves a requested version or alias by treating it as a filename under $NVM_DIR/alias. Before 0.40.8, nvm_alias() concatenated the requested name onto that directory and read the result with no containment check, so a name containing a `..` component escaped the alias directory; under the default layout an alias such as `../../.npmrc` resolves to a file in the user's home directory. nvm_print_alias_file() then emits every non-comment, non-empty line of whatever was opened. A version string taken from an untrusted .nvmrc reaches this path, so a developer who runs `nvm use`, `nvm install`, or `nvm which` inside an attacker-supplied repository discloses the first non-comment line of an arbitrary file readable by that user, in the resulting "is not yet installed" error message. A user-supplied `nvm alias <traversing-name>` discloses every non-comment line of the target file. There is no integrity or availability impact, and no command execution on this path.

vLLM through 0.29.0 contains a denial of service vulnerability in the NIXL connector's prefix caching implementation that fails to properly validate block counts across multi-prompt completion requests in prefill/decode disaggregated deployments. Attackers can trigger an assertion failure in NixlBaseConnectorWorker._apply_prefix_caching by submitting completion requests with multiple prompts of varying lengths, causing the decode worker to terminate and become unavailable until restarted.