Semaine du 5 octobre au 11 octobre 2026

zephyrproject zephyr

The zbus proxy agent IPC backend in subsys/zbus/proxy_agent/zbus_proxy_agent_ipc.c logged the channel name of a rejected inter-domain frame with a plain %s conversion. The frame type struct zbus_proxy_msg carries a fixed-size channel_name[] array as its last member, and nothing in the transport guarantees the array is NUL-terminated. The only code that verifies termination is zbus_proxy_agent_receive_cb() in subsys/zbus/proxy_agent/zbus_proxy_agent.c, which rejects the frame in precisely those cases — so the warning printed a non-terminated buffer exactly on the error paths where the name had been found invalid (or, for an invalid message_size, had not been inspected at all). Any peer domain able to place a frame of sizeof(struct zbus_proxy_msg) bytes on the bound ipc_service endpoint can trigger it, by sending a frame with an out-of-range message_size or with channel_name[] containing no NUL byte. Reaching the code requires CONFIG_ZBUS_PROXY_AGENT_IPC and logging built at warning level or above (the default), and requires control over the firmware of the peer domain — typically a second core on the same SoC. The resulting strlen() inside the log packager walks past the end of the frame object until it finds a zero byte. With the icmsg backend the frame lives in a stack buffer of the IPC work-queue thread, so bytes of that thread's stack are rendered into the log message; with the rpmsg backends the scan continues through the shared vring memory. Impact is bounded to disclosure of a small amount of adjacent memory into the receiving domain's log sink, plus a possible fatal fault if the scan leaves a mapped region; the log packager's own -ENOSPC bound prevents the overrun from becoming a write. The fix bounds the conversion with %.*s and MIN(msg->channel_name_len, sizeof(msg->channel_name)).

Hugging Face Text Embeddings Inference

Hugging Face Text Embeddings Inference through 1.9.4 contains a cleartext logging vulnerability that exposes the configured api_key because the router's Args struct lacks a redact attribute for it. Attackers with access to router logs, container output, or OTLP telemetry can recover the Bearer token and call the protected embedding and rerank endpoints.

thorsten phpMyFAQ

phpMyFAQ through 4.1.10 contains a missing authorization vulnerability in the MCP server faq_search tool that allows MCP clients to read restricted FAQs because Search::searchDatabase() never applies user or group permission checks. Attackers connected to the phpmyfaq:mcp:server can issue search queries to retrieve the full question and answer text of active FAQs restricted to specific users or groups.

tbphp gpt-load

GPT-Load through 1.4.11 contains a cleartext logging vulnerability that exposes client proxy keys because the Gin Logger middleware records the raw query string before extractAuthKey strips the key parameter. Attackers with read access to console logs or ./data/logs/app.log can recover proxy keys from Gemini-style requests and use them against the corresponding group.

DBeaver CloudBeaver

CloudBeaver through 25.3.5 contains a missing authorization vulnerability in WebSQLResultServlet that allows any web session holder to read other users' LOB export files from a shared folder. Attackers can guess table and column names and enumerate second-resolution timestamps to download victims' LOB values, including data from connections they cannot query.

shepherd-agents Shepherd

Shepherd (shepherd-ai) through 0.3.1 contains a server-side request forgery guard bypass in the citation-checker extra because the public_url guard validates a resolved address but fetch re-resolves the hostname at connect time. Attackers who plant a crafted reference URL in a checked document and control its DNS can rebind it to internal addresses, sending GET requests to internal HTTP(S) services and capturing responses in evidence files.

alexjustesen speedtest-tracker

Speedtest Tracker through 1.15.0 contains an IP allowlist bypass vulnerability that allows unauthenticated remote attackers to evade ALLOWED_IPS and Prometheus allowlists by spoofing X-Forwarded-For headers. Because bootstrap/app.php trusts every peer as a proxy, attackers can supply an allowlisted address to read /prometheus metrics and reach protected web and API endpoints.

QingYunA answer-me-with-html

answer-me-with-html through 0.5.0 contains a link following vulnerability in the am CLI code block src= embedding, where localPath() checks only path text without resolving symlinks. Attackers can ship a repository with a symlink pointing outside the checkout so am render embeds readable external files into generated HTML, disclosing them when shared.

InstantSoft icms2

A vulnerability has been found in InstantSoft icms2 up to 2.18.2. This issue affects the function index of the file templates/default/controllers/messages/index.tpl.php of the component Private Message Handler. Such manipulation of the argument nickname leads to cross site scripting. The attack can be launched remotely. The exploit has been disclosed to the public and may be used. The name of the patch is 3a1ec8fcb073a46d06a2ab83bc2bf68225834281. It is best practice to apply a patch to resolve this issue.

Unknown Dokan: AI Powered WooCommerce Multivendor Marketplace Solution

The Dokan: AI Powered WooCommerce Multivendor Marketplace Solution WordPress plugin before 5.2.0 does not verify that the vendor a commission calculation is requested for is the requesting vendor, allowing vendors to disclose the commission rate and fixed fee the marketplace administrator configured for other vendors.

Unknown Squadeno

The Squadeno WordPress plugin before 1.12.0 does not enforce its restrictions on every way a sport can be saved, allowing users with the lowest-tier Trainer role to change the section, age group, author, password, comment settings and date of a sport they are assigned to.

Unknown Envira Gallery

The Envira Gallery WordPress plugin before 1.16.2 does not verify that a user is authorized to read a gallery before rendering it, allowing authors to embed and expose other users' non-public gallery metadata to unauthenticated visitors.

Unknown Envira Gallery

The Envira Gallery WordPress plugin before 1.16.2 does not correctly check authorization on its gallery-conversion feature, verifying that the requester can edit an arbitrary post they name rather than that they are allowed to create the Envira Gallery WordPress plugin before 1.16.2's own gallery content, allowing users with contributor-level access to create and publish gallery posts that the Envira Gallery WordPress plugin before 1.16.2's settings otherwise withhold from them.

Unknown Envira Gallery

The Envira Gallery WordPress plugin before 1.16.2 does not verify that an image identifier added to a gallery refers to a media attachment the caller is permitted to view, allowing any user able to create and edit a gallery (Author and above by default) to disclose the title and excerpt of other users' private, draft, pending and trashed posts that WordPress would otherwise withhold from them.

addsumtech slide-maker

slide-maker through 5.8.0 contains a path traversal vulnerability in generate_images_openai.py that allows attackers to write image files outside the output directory via manifest-supplied filenames. Attackers can influence deck source material so the image prompt manifest contains ../ or symlinked filenames, creating directories and overwriting existing files at arbitrary paths.

mealie-recipes mealie

Mealie 3.26.0 through 3.28.0 contains a server-side request forgery vulnerability in the OpenID Connect avatar fetch that ignores ports when allowlisting the identity provider hostname. Authenticated OIDC users who control their picture URL can make the server send GET requests to arbitrary ports on the provider's internal address on each login.

ZTE Z80 Ultra

ZTE Z80 Ultra has a local information disclosure vulnerability. Third-party applications can capture data returned by system interfaces to obtain device-related information.

ZTE Z80 Ultra

ZTE Z80 Ultra has an interface permission validation vulnerability. The callable functions provided by the system lack sufficient access control. An attacker can leverage these functions to read relevant information.

ZTE Z80 Ultra

ZTE Z80 Ultra contains an information disclosure vulnerability, through which third-party applications can read relevant information by hooking system APIs.

softaculous SpeedyCache – Cache, Optimization, Performance

The SpeedyCache – Cache, Optimization, Performance plugin for WordPress is vulnerable to Sensitive Information Exposure in all versions up to, and including, 1.4.2 via the 'comment_author_*, comment_author_email_*' parameter. This makes it possible for unauthenticated attackers to extract the full name and email address of returning commenters pre-filled into comment form input fields and persisted as the site-wide cached page by any unauthenticated attacker requesting the same public URL. The read-side handler in advanced-cache.php correctly skips cached delivery for requests carrying comment_author_* cookies, but this check is absent on the write path, meaning the cache poisoning is invisible to the victim commenter yet fully exploitable by any unauthenticated attacker with no cookies.

Unknown Portfolio Filter Gallery

The Portfolio Filter Gallery WordPress plugin before 2.2.1 does not perform a per-object authorization check before deleting an attachment, allowing users with the Contributor role and above to permanently delete certain media attachments belonging to other users, including administrators.

Unknown Rank Math SEO

The Rank Math SEO WordPress plugin before 1.0.280 does not escape a stored redirection source value before outputting it in an administrative list view, allowing users who can manage redirections (Administrators by default) to store JavaScript that executes in the session of any user who later opens that view, including a Super Administrator on multisite.

senols AI Puffer – AI Chatbot, AI Writer & Automation

The AI Puffer – Chat. Create. Automate. (formerly AI Power) plugin for WordPress is vulnerable to authorization bypass in all versions up to, and including, 2.4.89. This is due to the plugin not properly verifying that a user is authorized to perform an action. This makes it possible for authenticated attackers, with subscriber-level access and above, to modify global site-wide semantic search settings, including vector provider, embedding provider, embedding model, target ID, number of results, and no-results text stored in aipkit_options, that are otherwise restricted to administrators. Exploitation requires that an administrator has previously granted the Knowledge Base ('sources') module to the attacker's role via the Role Manager, as this access is not available to lower-privileged users by default.

senols AI Puffer – AI Chatbot, AI Writer & Automation

The AI Puffer – Chat. Create. Automate. (formerly AI Power) plugin for WordPress is vulnerable to authorization bypass in all versions up to, and including, 2.4.89. This is due to the plugin not properly verifying that a user is authorized to perform an action. This makes it possible for authenticated attackers, with subscriber-level access and above, to modify global plugin indexing and vector-search configuration options (aipkit_training_general_settings and aipkit_indexing_field_settings), including chunking parameters, file upload visibility, and per-CPT field indexing settings, affecting all users of the plugin. This is only exploitable in configurations where an administrator has granted 'sources' module access to a lower-privileged role via the plugin's Role Manager.

contao

Contao is an Open Source CMS. From version 4.0.0 until 5.3.50 and 5.7.12, RequestTokenListener validates REQUEST_TOKEN only for POST requests, while the declarative GET guard runs only when an act parameter is present. Backend actions dispatched through the key parameter can therefore execute without a CSRF token when an authenticated backend user loads an attacker-controlled URL. Reachable actions remain limited to modules available to that user, and the advisory demonstrates destructive or state-changing actions rather than privilege escalation. This issue is fixed in versions 5.3.50 and 5.7.12.

illumos-gate · OmniOS

A missing authorization check in the illumos IP management daemon (ipmgmtd) allows a local user to change the persistent IP multipathing (IPMP) configuration. The ipmgmtd door dispatch table in usr/src/cmd/cmd-inet/lib/ipmgmtd/ipmgmt_door.c does not require the solaris.network.interface.config authorization for the IPMGMT_CMD_IPMP_UPDATE command, although its handler, ipmgmt_ipmp_update_handler(), writes to the persistent ipadm configuration when the IPMGMT_PERSIST flag is set. An unprivileged local user can therefore add interfaces to, or remove them from, existing IPMP groups in the stored configuration. The running configuration is not changed; the modification takes effect when the stored configuration is next applied, such as at boot, and may disrupt network connectivity. The flaw has existed since 2021 (illumos-gate commit a73be61a), and affects any illumos distribution prior to illumos-gate commit e8d3efa1.

illumos-gate · OmniOS

A missing authorization check in the illumos zones statistics daemon (zonestatd) allows a local user in any zone to disrupt zonestat in other zones and to determine which zones are running. The zonestatd door server procedure, zsd_server() in usr/src/cmd/zonestat/zonestatd/zonestatd.c, handles the ZSD_CMD_NEW_ZONE command, which is intended to be sent by zoneadmd, without checking the caller's credentials. Because the zonestatd door is accessible to all users in every zone, an unprivileged user can send this command with an arbitrary zone ID, causing zonestatd to re-create its door file in that zone, so that new zonestat requests in that zone can fail while the file is replaced. The time taken to handle the command also reveals whether a given zone ID belongs to a running zone. The flaw has existed since 2010 (illumos-gate commit efd4c9b6), and affects any illumos distribution prior to illumos-gate commit 865b58d2.

canva affinity

The Affinity by Canva app for macOS before 3.3.1 (October 2026 release) did not perform adequate bounds checking when generating QuickLook thumbnails and previews of Affinity document files, leading to an out-of-bounds heap read. A threat actor could craft an Affinity document that, when displayed or previewed by a user in Finder, could disclose the contents of adjacent heap memory, including memory addresses, in the rendered thumbnail or preview image.

canva affinity

The Affinity by Canva application for macOS before 3.3.1 (October 2026 release) did not validate image dimensions against the size of the pixel data when generating QuickLook thumbnails and previews of Affinity document files, leading to an out-of-bounds heap read. A threat actor could craft an Affinity document that, when displayed or previewed by a user in Finder, could disclose the contents of adjacent heap memory in the rendered thumbnail or preview image, or cause the thumbnail or preview extension to crash.

canva affinity

The Affinity by Canva app before 3.3.1 (October 2026 release) did not perform adequate bounds checking when parsing Affinity document files, leading to an out-of-bounds pointer dereference. A threat actor could craft an Affinity document that, when opened by a user in Affinity, could result in an application crash.

canva affinity

The Affinity by Canva application before 3.3.1 (October 2026 release) did not perform adequate bounds checking when parsing arrays of strings in Affinity document files, leading to a heap buffer over-read. A threat actor could craft an Affinity document that, when opened by a user in Affinity, could disclose the contents of adjacent heap memory in the document's text or result in an application crash.

canva affinity

The Affinity by Canva application before 3.3.1 (October 2026 release) did not correctly handle incomplete UTF-8 character sequences when parsing text in Affinity document files, leading to a heap buffer over-read. A threat actor could craft an Affinity document that, when opened by a user in Affinity, could disclose the contents of adjacent heap memory in the document's text or result in an application crash.

Dell Secure Connect Gateway (SCG) Policy Manager · Dell policy manager for secure connect gateway

Dell Secure Connect Gateway (SCG) Policy Manager, versions prior to 5.34.00.16, contains an Improper Check for Unusual or Exceptional Conditions vulnerability. A high privileged attacker with remote access could potentially exploit this vulnerability, leading to Information tampering and Protection mechanism bypass.

Unknown Code Snippets

The Code Snippets WordPress plugin before 3.10.0 does not perform a capability check on one of its snippet-management actions and derives the network scope of the targeted snippet from the request instead of from the stored record, allowing an administrator of a single subsite on a multisite network to activate, deactivate and reprioritise network-scoped snippets that run across every site in the network.

strawberry-graphql strawberry

Strawberry GraphQL is a library for creating GraphQL APIs. From 0.312.3 until 0.327.2, the legacy graphql-ws subscription handler in strawberry/subscriptions/protocols/graphql_ws/handlers.py does not remove naturally completed operations from self.tasks and self.subscriptions. When max_subscriptions_per_connection is configured, a client on a persistent WebSocket connection can use distinct operation IDs for one-shot subscriptions to fill the connection's configured slots even after those subscriptions send complete, causing later legitimate operations on that connection to be rejected with Subscription limit reached. The modern graphql-transport-ws protocol and deployments without the configured per-connection cap are not affected. This issue is fixed in version 0.327.2.

Incomplete X.509 implementation in GnuTLS v3.8.13 allows attackers controlling a subordinate Certificate Authority to bypass cross-domain PKI restrictions and issue unauthorized certificates.

An issue in gnutls v.3.8.13 allows an attacker to obtain sensitive information via failing to reject end-entity X.509 certificates that contain a contradictory combination of Key Usage (KU) and Extended Key Usage (EKU)

MongoDB C Driver

An out-of-bounds read in SCRAM authentication response parsing in the MongoDB C Driver can read one byte beyond a fixed-size buffer when processing a malformed server-final message. A server or network intermediary able to provide this message before server-signature verification can cause the application using the driver to terminate. The extra byte is not returned through the protocol.

mcollina msgpack5

msgpack5 is a msgpack v5 implementation for node.js and the browser. Prior to 6.1.0, decoding a negative signed 64-bit integer modifies the corresponding bytes in the caller-provided input buffer while computing the value. Applications that retain or reuse encoded input for integrity checks, logging, or later processing can observe silently corrupted data, while positive integers and other MessagePack value types are unaffected. This issue is fixed in version 6.1.0.

FFmpeg · Ffmpeg ffmpeg

FFmpeg through 9.0.2 contains an uninitialized memory disclosure vulnerability in av_dynamic_hdr_plus_to_t35() that leaves up to three payload bytes uninitialized when tone_mapping_flag is 0. Attackers can supply crafted Matroska T.35 BlockAdditional or HEVC/AV1 SEI metadata so that remuxing or transcoding writes leaked process memory into output files.

pydantic-ai · pydantic-ai-slim · Pydantic pydantic ai

Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. From 1.77.0 until 1.107.6 and 2.44.0, the local web_fetch_tool and the WebFetch local fallback compare blocked_domains entries with a URL hostname before both values are normalized to the form used by getaddrinfo. An attacker-influenced model can use an equivalent IDNA spelling, non-ASCII label separator, case variation, or trailing root label that resolves to a blocked host but does not match the configured string, causing the application to fetch that host with its own privileges. allowed_domains fails closed for unmatched spellings, and private-IP and cloud-metadata protections remain effective. This issue is fixed in versions 1.107.6 and 2.44.0.

İzometri IT Services Domestic and Foreign Trade Co. Ltd. eimzamip

Improper neutralization of Script-Related HTML tags in a web page (basic XSS) vulnerability in İzometri IT Services Domestic and Foreign Trade Co. Ltd. Eimzamip allows Stored XSS. This issue affects eimzamip: from v1.6.4 before v1.6.7.

ScriptKittyOS beam_mcp

Incorrect Type Conversion or Cast vulnerability in BeamMCP.Server in ScriptKittyOS beam_mcp allows an MCP client's JSON true, false and null tool arguments to reach the host's dispatch function as the strings "true", "false" and "nil". After BeamMCP.Schema.validate/2 accepted a value as a boolean, normalize_arguments/2 passed every argument through to_json_value/1, whose atom clause converts true, false and nil to strings. A string is truthy in Elixir, so a host that tests a boolean argument, for example if args.dry_run, takes the opposite branch for false, and a guard such as confirm: false reads as set. The client controls the argument and could send true directly, so the practical impact is limited to hosts whose behaviour on false differs from their behaviour on true, and to any policy layer in front of the server that permits false but refuses true. The same normalisation applies to prompts/get arguments, which exist from 0.5.0. This issue affects beam_mcp: from 0.1.0 before 0.10.1.

mutt

heap OOB write in convert_file_from_to() via a crafted Content-Type header allows attacker to OOB write when email is used as a template.