Vulnerabilities exploitable today
373,979in current view
Single score combining CVSS, KEV membership and EPSS. Every CVE with its own record — timeline from publication to active exploitation.
In KEV catalog1,709
New KEV · 24H0
Exploit Today ≥ 701,644
Distribution · last window
- Critical2,324
- High8,430
- Medium6,377
- Low704
Window
Severity
Flags
CVECVSSEPSSKEVRExploitTitleMod.
CVE-2026-575797.5 HIG—
———Alchemy is an open source content management system engine written in Ruby on Rails. Prior to 7.4.15, 8.0.15, 8.1.14, and 8.2.6, the unauthenticated GET /api/pages/nested endpoint implemented by Api::PagesController#nested in app/controllers/alchemy/api/pages_controller.rb returns an unfiltered page tree because it performs no authorization and does not scope descendants by the caller's ability. Anonymous callers can retrieve restricted and unpublished page metadata that the sibling show action denies. When elements=true is supplied, PageTreeSerializer also returns element and ingredient content from restricted pages because PageTreePreloader and the serializer do not apply an ability check to those records. This issue is fixed in versions 7.4.15, 8.0.15, 8.1.14, and 8.2.6.18hCVE-2026-34151——
———XWiki Platform is a generic wiki platform. Prior to 17.10.5 and 18.2.0, the /skin/ action in com.xpn.xwiki.web.SkinAction can resolve double-encoded parent-directory segments outside the intended skin or web-application resource prefix when Jetty 12 or later decodes the request path. The affected lookup is replaced with Environment.getResourceAsStream(String, String), which constrains a resource to its expected prefix. An unauthenticated remote attacker can use the vulnerable behavior to read arbitrary resources permitted to the Jetty process, including WEB-INF/xwiki.cfg and, depending on deployment depth and operating-system permissions, host files. Tomcat and Jetty versions before 12 do not appear affected. This issue is fixed in versions 17.10.5 and 18.2.0.17hCVE-2026-571156.5 MED—
———PraisonAI is a multi-agent teams system. Prior to praisonaiagents 1.6.59, SpiderTools.scrape_page validates only the initial URL and lets requests.Session.get follow redirects automatically, so a public-looking URL can redirect to a loopback, private, link-local, or metadata address without revalidation. The redirected response body is returned through scrape_page and its extract_links, crawl, and extract_text callers, allowing disclosure from otherwise blocked services. This issue is fixed in praisonaiagents 1.6.59.17hCVE-2026-256878.1 HIG—
———A race condition in the ZPA tunnel handler of affected versions of Zscaler Client Connector (ZCC) allows a heap corruption, resulting in a denial of service (client crash) and potentially arbitrary code execution in the context of the ZCC process.8hCVE-2026-15600——
———Alior Bank PrestaShop module "raty" for commercial partners is vulnerable to SQL Injection in the toggleCategoryPromotionAction method. The module inserts value of the POST parameter "status" into SQL UPDATE queries without any sanitization or validation. An attacker with access to the product or category add/edit functionality in the PrestaShop backoffice can inject arbitrary SQL, potentially allowing unauthorized access to and modification of database contents.16hCVE-2026-129856.8 MED—
———Mattermost versions 11.9.x <= 11.9.0, 11.8.x <= 11.8.4, 11.7.x <= 11.7.7 Mattermost failed to validate Dynamic Client Registration redirect URIs by URL component (matching glob patterns against the raw URI string instead) which allows a remote unauthenticated attacker to register an OAuth client with an attacker-controlled callback host that bypasses the configured redirect URI allowlist via a crafted redirect URI that places an allowlisted host/path suffix inside the query string.. Mattermost Advisory ID: MMSA-2026-007008hCVE-2026-472565.3 MED—
———OpenTelemetry, also known as OTel, is a vendor-neutral open source Observability framework for instrumenting, generating, collecting, and exporting telemetry data such as traces, metrics, and logs. Prior to 0.154.0, the Sentry exporter reads the remote OTLP sender-controlled service.name resource attribute in exporter/sentryexporter/sentry_exporter.go through extractProjectSlug and getOrCreateProjectEndpoint, passes the raw project slug to GetOTLPEndpoints and GetProjectKeys in exporter/sentryexporter/sentry_client.go, and interpolates it into a Sentry API URL without applying projectSlugRegexp through validateRoutingConfig at runtime in exporter/sentryexporter/config.go. Special characters can turn the expected path suffix into query data in all deployments or introduce slash and dot segments that traverse paths when the Sentry deployment normalizes them, while the Collector attaches its operator-configured bearer token to the request. A successful request can reach token-authorized administrative, organization, member, or key endpoints within the configured Sentry organization, and an attacker-controlled project slug can redirect subsequently exported telemetry. Sentry token middleware prevents cross-organization access. This issue is fixed in version 0.154.0.17hCVE-2026-909497.8 HIG—
———A flaw was found in GIMP's PSP (Paint Shop Pro) file loader. When processing a compressed selection channel, a heap-based buffer overflow can occur due to a mismatch between the allocated buffer size and the amount of data decompressed. A remote attacker could exploit this vulnerability by crafting a malicious PSP file. Opening this file in GIMP could lead to a crash or arbitrary code execution.20hCVE-2026-909487.8 HIG—
———A flaw was found in GIMP's ICO file loader. When processing an ICO file containing an embedded PNG image, an integer overflow can occur during the calculation of the required buffer size. This leads to an undersized buffer being allocated, causing a heap-based buffer overflow when the decoded pixel data is written. A remote attacker could exploit this by crafting a malicious ICO file, which, when opened, could lead to arbitrary code execution or a crash.20hCVE-2026-909405.3 MED—
———novel-plus through 5.3.3 contains an insecure default cache-management password in the CacheController.refreshCache endpoint that allows anonymous attackers to invalidate portal caches by supplying the hardcoded default value in the URL path. Attackers can trigger unauthorized cache invalidation by accessing the cache/refresh endpoint with the known default password, forcing unnecessary database queries to repopulate the cache.22hCVE-2026-494003.3 LOW—
———October System provides the system module for October Content Management System. Prior to versions 3.7.17 and 4.2.21, the backend `SessionMaker` trait stored widget session state as `base64(serialize(...))` and consumed it with `unserialize()` without an `allowed_classes` restriction. Any code path that could write to a `widget.*` session key with attacker-controlled bytes could trigger PHP object injection the next time the widget read its session state, allowing instantiation of arbitrary classes and reachable PHP gadget chains. This issue only affects installations running with `cms.safe_mode` enabled. Safe Mode is a niche opt-in feature, primarily used for demo installations and multi-tenant or shared-editor scenarios where untrusted users are deliberately granted access to the CMS markup editor. In standard production deployments Safe Mode is off, backend access is restricted to trusted administrators, and a markup editor can already execute arbitrary PHP directly. The session-write path that reaches this sink is gated by the Safe Mode sandbox, so installations without Safe Mode enabled are not exposed. Scope of impact is narrow even with Safe Mode enabled. The standard backend code paths that populate widget session state (search terms, sort options, selected IDs, filter values) wrap the input inside a known array shape before serializing, so user-supplied values never reach `unserialize()` as a controllable serialized payload. Exploitation requires the Safe Mode session-write path together with a suitable PHP gadget chain reachable from the installed dependency set. The hardening below removes the underlying object-injection sink so the class of issue is closed off regardless. The vulnerability has been patched in v3.7.17 and v4.2.21. Two changes were applied. `Backend\Traits\SessionMaker` now stores widget session state as plain JSON instead of `base64(serialize(...))`, eliminating the object-injection sink entirely for new writes. Reads transparently fall back to the legacy format for one upgrade cycle so existing sessions retain their saved widget state. The legacy `unserialize()` fallback path now sets `allowed_classes => false`, so even values written before the upgrade cannot instantiate objects. As a workaround, restrict CMS markup editing access to fully trusted administrators only, the standard October CMS recommendation for any deployment.18hCVE-2026-909396.5 MED—
———novel-plus through 5.3.3 contains an information disclosure vulnerability in the /sys/user/list endpoint that lacks proper permission annotations. Authenticated attackers can retrieve password hashes and personal data including email addresses and phone numbers for users within their data scope, enabling offline hash cracking and account takeover.21hCVE-2026-53495——
———containerd is an open-source container runtime. Prior to 1.7.35, 2.0.12, 2.2.8, and 2.3.5, containerd on Linux with the CRI plugin enabled can indefinitely block the drainExecSyncIO goroutine in internal/cri/server/container_execsync.go when CRI ExecSync is used by exec probes or lifecycle hooks that launch long-lived background child processes retaining standard input and output pipes. The input and output drain phase has no default timeout and did not stop when the request context was canceled, so repeated ExecSync invocations can accumulate blocked goroutines and host memory. The resulting resource exhaustion can cause the OOM killer to terminate containerd, leaving the container runtime unavailable until restart. Deployments not using containerd's CRI implementation and containers not running on Linux are not affected. This issue is fixed in versions 1.7.35, 2.0.12, 2.2.8, and 2.3.5.18hCVE-2026-541788.1 HIG—
———backpack/crud provides Create, Read, Update & Delete (CRUD) functions for Backpack, a collection of Laravel packages that help users build custom administration panels. Prior to 6.8.12 and 7.0.35, HasUploadFields::uploadMultipleFilesToDisk in src/app/Models/Traits/HasUploadFields.php trusts disk-relative paths from clear_<attribute>[] and passes them to Storage::disk()->delete without confirming that the paths are persisted on the current model record. An authenticated user with Update access to a CRUD using this mutator through src/app/Models/Traits/CrudTrait.php can delete another record's attachment, a shared asset, or another operational file on the configured disk by submitting its path. The newer MultipleFiles uploader is not affected because it intersects requested deletions with the record's persisted file list. This flaw does not permit reading the deleted files. The 5.x line remains affected through its final releases. This issue is fixed in versions 6.8.12 and 7.0.35.18hCVE-2026-541828.1 HIG—
———backpack/crud provides Create, Read, Update & Delete (CRUD) functions for Backpack, a collection of Laravel packages that help users build custom administration panels. Prior to 4.1.70, 5.6.2, 6.8.13, and 7.0.36, Backpack\CRUD\Stats::makeCurlRequest in src/Stats.php is reached from BackpackServiceProvider::boot() and constructs a shell command with a URL influenced by the HTTP Host header, which it passes to exec() without adequate shell neutralization. An unauthenticated attacker whose malformed Host value reaches PHP can inject operating-system commands when exec() and curl are available and the 1-in-100 random gate is reached. Repeated requests can reach the random gate. Successful exploitation executes commands as the web-server user, exposing environment secrets, files, and reachable services and permitting data modification or service disruption. Common reverse-proxy Host validation and hardened PHP configurations that disable exec() reduce reachability but do not correct the vulnerable construction. This issue is fixed in versions 4.1.70, 5.6.2, 6.8.13, and 7.0.36.17hCVE-2026-558476.1 MED—
———Allure 2 is the version 2.x branch of Allure Report, a multi-language test reporting tool. Prior to 2.39.0, the ansi.js helper at allure-generator/src/main/javascript/helpers/ansi.js passes attacker-influenced statusMessage and statusTrace values through AnsiToHtml without HTML escaping and wraps the result in Handlebars SafeString, disabling template auto-escaping in allure-generator/src/main/javascript/blocks/status-details/status-details.hbs. JunitXmlPlugin.java can populate these fields directly from crafted JUnit XML failure messages and traces, and equivalent input flows exist in the TRX, xUnit XML, xctest, and Allure1 and Allure2 plugins. When a user views the affected status details, unescaped markup executes arbitrary JavaScript in the report origin, which can expose report data and compromise sessions associated with that origin. This is an incomplete-fix case because PR 3271 escaped link helpers but did not address the ANSI helper. This issue is fixed in version 2.39.0.16hCVE-2026-57577——
———DotVVM is an open source MVVM framework for web applications. Prior to 4.2.11, 4.3.15, and 5.0.0-preview09-final, a route containing multiple unconstrained parameters in one path segment can cause excessive regular-expression backtracking in DotvvmRoute.IsMatch when a remote requester supplies a long near-match path. DotvvmRouteParser.RouteRegex previously had no matching timeout. Patched runtimes retry with the .NET non-backtracking engine, while runtimes that do not support non-backtracking matching return HTTP 503 after the one-second timeout in DotvvmRoutingMiddleware. This issue is fixed in versions 4.2.11, 4.3.15, and 5.0.0-preview09-final.18hCVE-2026-907884.7 MED—
———A security flaw has been discovered in magicblack MacCMS10 2026.1000.4055. Affected by this vulnerability is an unknown functionality of the file /admin1.php/admin/template/index/path/.%40template%40default%40html%40label.html of the component Template Handler. Performing a manipulation results in os command injection. The attack may be initiated remotely. The exploit has been released to the public and may be used for attacks. The project was informed of the problem early through an issue report but has not responded yet.15hCVE-2026-575815.3 MED—
———DotVVM is an open source MVVM framework for web applications. Prior to 4.2.11, 4.3.15, and 5.0.0-preview09-final, applications with configured file upload storage allow unauthenticated users to submit files directly to DotvvmFileUploadMiddleware without an X-DotVVM-UploadToken generated by the FileUpload component. An attacker can repeatedly upload files and fill application storage, causing denial of service. DotvvmConfiguration.Security.AuthorizeFileUpload can additionally restrict which users may upload files. This issue is fixed in versions 4.2.11, 4.3.15, and 5.0.0-preview09-final.18hCVE-2026-907877.3 HIG—
———A vulnerability was identified in Soarkey StudentManagement up to e08f7f1d5015af407aa4cca0ada3dea189b4937e. Affected is the function RegisterServlet.doPost of the file code/WebContent/register.html of the component Registration Workflow. Such manipulation of the argument level leads to improper privilege management. The attack can be launched remotely. The exploit is publicly available and might be used. This product does not use versioning. This is why information about affected and unaffected releases are unavailable. The project was informed of the problem early through an issue report but has not responded yet.15hCVE-2026-907865.3 MED—
———A vulnerability was determined in Dvidelabs flatcc up to 0.6.3. This impacts the function align_order_members of the file src/compiler/semantics.c of the component Duplicate Symbol Handler. This manipulation causes reachable assertion. The attack can be initiated remotely. The exploit has been publicly disclosed and may be utilized. Patch name: 8b19ba4e992ebcad7f5970704d1afc5507fa5205. To fix this issue, it is recommended to deploy a patch.15hCVE-2026-734947.4 HIG—
———blaze is a Scala library for building asynchronous pipelines, with a focus on network IO. Prior to 0.23.18 and from 1.0.0-M1 until 1.0.0-M42, five HTTP/1.1 conformance laxities in the hand-written Java parser under http/src/main/java/org/http4s/blaze/http/parser/ can cause blaze to derive a different request boundary than a stricter fronting intermediary. A default BlazeServerBuilder accepts invalid or valueless header field names that violate tchar syntax, obsolete folded field lines (obs-fold), unsupported Transfer-Encoding values, duplicate Content-Length fields, and requests containing both Transfer-Encoding and Content-Length. If a lenient or legacy proxy forwards the malformed bytes but interprets them differently, the disagreement can permit front-end authorization bypass, response-queue poisoning on pooled backend connections, or cache poisoning. Exploitation requires a pair of disagreeing parsers; no non-default blaze configuration is required. The affected checks are enforced in BodyAndHeaderParser and Http1ServerParser. This issue is fixed in versions 0.23.18 and 1.0.0-M42.16hCVE-2026-86836——
———In Eclipse Ankaios versions 0.1.0 through 1.0.2, the agent creates workload files and Control Interface named pipes (FIFOs) under a predictable path derived from the agent name and a hash of the workload's runtime configuration. If a directory or FIFO already exists at that path when the agent (re)starts, the agent reuses it based only on an existence and/or file-type check, without validating its owner or permissions. A local, unprivileged user with write access to the same base directory (by default under `$TMPDIR/ankaios`, e.g. shared `/tmp`) can pre-create this path hierarchy, including the two Control Interface FIFOs, before the agent starts. The agent then treats the attacker-owned FIFOs as the legitimate Control Interface for the targeted workload. The attacker can complete the Control Interface handshake and issue requests using that workload's configured `controlInterfaceAccess` permissions, allowing impersonation of the workload and, depending on its configured permissions, unauthorized reading and/or modification of the cluster's desired state.17hCVE-2026-158936.5 MED—
———net_if_ipv6_calc_reachable_time() in subsys/net/ip/net_if.c derives a randomized ND reachable time from ipv6->base_reachable_time as min_reachable + sys_rand32_get() % (max_reachable - min_reachable), where min_reachable = base/2 and max_reachable = 3*base/2 using integer division. When base_reachable_time is 1, both min_reachable and the modulus collapse so the function returns 0, and net_if_ipv6_set_reachable_time() stores that 0 into ipv6->reachable_time.
The base_reachable_time is attacker-controlled: handle_ra_input() in subsys/net/ip/ipv6_nbr.c accepts the Reachable Time field of an incoming Router Advertisement whenever it is nonzero and <= MAX_REACHABLE_TIME, so a single unauthenticated, link-local RA carrying a Reachable Time of 1 drives the computed reachable time to 0. Router Advertisements are unauthenticated by default and require only adjacency to the target link.
When a neighbor is subsequently confirmed reachable, net_ipv6_nbr_set_reachable_timer() reads the value and executes NET_ASSERT(time, "Zero reachable timeout!"). On builds with CONFIG_ASSERT enabled this triggers a fatal kernel assertion — a remote denial of service; on builds without assertions the reachable timer is armed with K_MSEC(0) and fires immediately, forcing reachable neighbors into perpetual re-solicitation (STALE), degrading Neighbor Discovery. The impact is limited to availability; there is no memory-safety, confidentiality, or integrity consequence.15hCVE-2026-868307.2 HIG—
———Incorrect privilege assignment in Temporary Elevated Access Management (TEAM) for AWS IAM Identity Center solution before version 1.5.1 might allow an authenticated remote user with application-level access to read, approve, modify, or revoke arbitrary access requests, thereby obtaining unintended temporary elevated access to the AWS accounts accessed using the TEAM deployment.
This issue has been addressed in TEAM version 1.5.1 or later. We recommend upgrading to the latest version and ensuring any forked or derivative code is patched to incorporate the new fixes.15hCVE-2026-907855.3 MED—
———A vulnerability was found in Dvidelabs flatcc up to 0.6.3. This affects the function analyze_struct of the file src/compiler/semantics.c of the component Struct Analysis. The manipulation results in reachable assertion. It is possible to launch the attack remotely. The exploit has been made public and could be used. The patch is identified as f705032346ee39efd7d3848c50b73d455d28d06d. A patch should be applied to remediate this issue.15hCVE-2026-890216.9 MED—
———MikroTik RouterOS before 7.24.2 contains a path traversal vulnerability in the container package OCI/tar image extraction that allows attackers to write files outside the container root by supplying a crafted container image with symlinks pointing to arbitrary paths. Attackers can exploit unsanitized tar member path extraction during container import via /container/add to achieve root-privileged file creation, directory creation, file deletion via overlayfs whiteout, and hardlink creation on the persistent data partition without ever starting the container. The 7.23.x long-term branch does not contain this fix; the container binaries in container-7.23.3.npk and container-7.23.4.npk are byte-identical, and there is no fixed long-term release at the time of publication.17hCVE-2026-863494.3 MED—
———Mattermost versions 11.9.x <= 11.9.0, 11.8.x <= 11.8.4, 11.7.x <= 11.7.8, 10.11.x <= 10.11.22 fail to limit the nesting depth in the server-side Markdown parser which allows an authenticated attacker to cause a denial of service (CPU resource exhaustion) via a crafted post containing deeply nested blockquotes or list items.. Mattermost Advisory ID: MMSA-2026-0070716hCVE-2026-863484.3 MED—
———Mattermost versions <=11.9 11.0.9 11.4.8 11.7.7 10.22.11.0 fail to recover from handler panics, which allows an authenticated user to crash the plugin via a post-action request with an unexpected field type.. Mattermost Advisory ID: MMSA-2026-0070116hCVE-2026-159245.9 MED—
———Zephyr's TLS socket layer in subsys/net/lib/sockets/sockets_tls.c keeps a single process-global array, client_cache, of cached client sessions that is shared by every TLS socket context. The functions that mutate and read it — tls_session_save(), tls_session_get(), tls_session_cache_reset(), and the settings restore handler — allocate, free, and dereference each entry's heap buffer (entry->session). Before the fix these accesses were serialized only by the per-socket context mutex ctx->lock (assigned per socket in ctx_set_lock()), which provides no mutual exclusion between different sockets touching the shared cache.
Because CONFIG_NET_SOCKETS_TLS_MAX_CLIENT_SESSION_COUNT defaults to 1, any two concurrent client sockets contend for the same slot. A thread in tls_session_get() reading entry->session inside mbedtls_ssl_session_load() can run concurrently with another thread in tls_session_save() that selects the same entry for reuse and executes mbedtls_free(entry->session) before reallocating — a use-after-free read, and a double-free when two saves evict the same entry. Both corrupt the mbedTLS heap. The cache is reached on ordinary client paths: at connect time via tls_session_store()/tls_session_restore(), and (on main) whenever a TLS 1.3 session ticket arrives during recv()/poll() via tls_session_store_current().
Exploitation requires an application that opts into per-socket client session caching (the TLS_SESSION_CACHE socket option, off by default) and runs concurrent TLS client connections on multiple threads; the timing that opens the window is influenced by the remote peer(s), so a malicious or compromised server can raise session-ticket frequency to widen it. The reliably-demonstrable impact is memory corruption leading to a crash or heap corruption (denial of service). The fix adds a dedicated session_cache_lock mutex taken across every accessor of client_cache, serializing all reads and frees and closing the race.15hCVE-2026-161476.8 MED—
———The ITE IT82xx2 USB device-controller driver (drivers/usb/udc/udc_it82xx2.c) mishandles multi-packet OUT transfers on non-control endpoints. In work_handler_out() the active transfer buffer is obtained with udc_buf_peek() (which does not dequeue it); when a full max-packet-size packet arrives but the buffer still has tailroom (the transfer is not yet complete), the pre-fix code both re-arms the endpoint to keep filling that same buf via work_handler_xfer_continue() and simultaneously hands the same, still-being-filled buffer to the upper stack with udc_submit_ep_event().
Because udc_submit_ep_event() transfers ownership of the buffer to the USB device stack (usbd_event_carrier() appends &buf->node to uds_ctx->ep_events, after which the class handler processes and net_buf_unref()s it), the driver continues to DMA subsequent host-controlled OUT packets into a buffer the upper stack may already have freed and recycled — a use-after-free write. In addition, since the buffer was never dequeued, the completing packet runs udc_buf_get() on the same object and submits it a second time, appending &buf->node to the event slist twice (singly-linked-list corruption) and causing a double net_buf_unref().
The IT82xx2 is a USB peripheral controller, so the untrusted USB host controls OUT-transfer packetization and can force this path against any non-control OUT endpoint whose queued buffer exceeds one packet — an ordinary bulk/interrupt pattern. The driver and USB device stack run in kernel context above the external host, giving the host a device-side kernel heap-corruption primitive: a reliable denial of service and, because the written bytes are attacker-controlled, plausible corruption of adjacent net_buf pool memory. The vector is physical (USB attach). The fix defers submission until the buffer is completely filled and lets xfer_work_handler() drive continuation, so each OUT buffer is submitted to the upper stack exactly once.15hCVE-2026-841796.5 MED—
———Description
getTopologyPageInfo merged the Nimbus daemon configuration with the topology's own configuration and returned the result without redaction in the topology_conf field of TopologyPageInfo. The Storm UI copied that value verbatim into the configuration field of GET /api/v1/topology/{id} and of the corresponding metrics endpoint.
Where the cluster is configured with them, the merged map includes storm.zookeeper.auth.payload, which Storm's own documentation directs operators to keep in storm-cluster-auth.yaml under permissions that deny access from workers, together with the keystore and truststore passwords for the Thrift, Netty and ZooKeeper TLS configuration, and any plugin key whose name denotes a secret.
getTopologyPageInfo is a topology read-only operation. Under SimpleACLAuthorizer a principal listed in topology.readonly.users or topology.readonly.groups could therefore read daemon credentials that the dedicated cluster configuration API, getNimbusConf, redacts and that is gated on nimbus.users instead. The sibling operations that exist to serve configuration were masked; the topology page, which merges in strictly more daemon state, was not.
Mitigation
Upgrade to 3.1.0, where credential-bearing values are masked before any configuration is served over the Nimbus API.
Users who cannot upgrade immediately should remove any principal that is not trusted with cluster credentials from topology.readonly.users, topology.readonly.groups, topology.users and topology.groups, and should rotate the ZooKeeper authentication payload and any TLS keystore or truststore passwords that were reachable through the topology page.
Credit
Wanxin Yin (yaklang.io) reported this issue to the Apache Security Team.15hCVE-2026-537176.5 MED—
———Envoy Gateway is an open source project for managing Envoy Proxy as a standalone or Kubernetes-based application gateway. Prior to 1.7.4 and 1.8.1, internal/wasm/imagefetcher.go follows tenant-controlled EnvoyExtensionPolicy spec.wasm[].code.image.url values to Docker or OCI Wasm layers, and extractWasmPluginBinary uses the untrusted tar-header h.Size value to allocate memory before validating the entry name or declared size. A small PAX or GNU tar header can therefore claim a multi-terabyte entry even though the surrounding LimitReader restricts only the bytes read from the stream, and no registry allowlist prevents a permitted tenant from selecting an attacker-controlled registry that the controller can reach. The allocation is attempted for every tar entry and can cause an unrecoverable Go runtime out-of-memory failure; because the custom resource persists, reconciliation repeatedly crash-loops the shared controller and causes a single-request, non-volumetric, cluster-wide control-plane denial of service. This issue is fixed in versions 1.7.4 and 1.8.1.16hCVE-2026-542474.3 MED—
———Skipper is an HTTP router and reverse proxy for service composition. Prior to 0.26.22, Handler in dataclients/kubernetes/admission/admission.go passes the body of requests to the Kubernetes admission endpoint at :9443/admission directly to io.ReadAll(r.Body) without a size limit. An attacker with in-cluster network access and a valid Kubernetes client certificate can send a very large body that causes unbounded memory allocation and an out-of-memory termination of the Skipper process. The disruption is limited to Ingress and RouteGroup admission rather than pod creation or unrelated admission controllers, and Kubernetes normally restarts the process. This issue is fixed in version 0.26.22.16hCVE-2026-824399.8 CRI—
———Description
The DRPC server kept a map from function name to request queue and created an entry the first time a
function name was seen. No code path ever removed an entry: request cleanup removed the request from its
queue, and the shutdown path drained queues, but the queue object and its map entry remained for the life of
the process.
Function names come from the client and are not constrained to functions any topology has registered, so the
number of retained entries is bounded only by the number of distinct names an attacker chooses to send, and
each retained entry holds the name itself. `drpc.authorizer` is unset by default, so no credentials are
required to reach the endpoint.
The retained state is permanent rather than a transient load spike, so the effect accumulates until the DRPC
server exhausts its heap.
Mitigation
Upgrade to 3.1.0, where a function's queue is removed once nothing is waiting in it.
Users who cannot upgrade immediately should configure `drpc.authorizer` so that only trusted principals can
reach the DRPC endpoints, and should ensure the DRPC ports are not reachable from untrusted networks.
Credit
The ASF -- found using Claude agents to study the security of open-source projects, validated and reported by Apache Storm.15hCVE-2026-544478.4 HIG—
———garminconnect is a Python 3 API wrapper for Garmin Connect that retrieves statistics and manages activities. Prior to 0.3.5, garminconnect/client.py Client.dump creates the OAuth token directory and garmin_tokens.json without explicit owner-only modes, so a permissive umask such as 022 can leave the directory mode at 0755 and the token file mode at 0644. garmin_tokens.json contains di_refresh_token, and another unprivileged user on a shared Linux or macOS host can read the token and obtain persistent access to the victim's Garmin Connect account, including health, fitness, activity, and device data. The Garmin.login tokenstore path is affected, and a pre-existing loosely permissioned token file remains exposed until rewritten or manually restricted. This issue is fixed in version 0.3.5.16hCVE-2026-550936.1 MED—
———Tract is a tiny, no-nonsense, self-contained TensorFlow and ONNX inference toolkit. Prior to 0.21.16, 0.22.2, and 0.23.1, tract-nnef uses unchecked usize multiplication in nnef/src/tensors.rs read_tensor for attacker-controlled tensor dimensions, the allocation size, and the reported tensor length. Loading a crafted NNEF archive through model_for_path or model_for_read reaches the default DatLoader and can make the wrapped size check accept a small allocation while data/src/tensor.rs as_slice_unchecked creates a much larger logical slice. Model construction through as_uniform can then read beyond the heap allocation and disclose adjacent data, and later access can terminate the process with a segmentation fault. The affected dense numeric tensor path does not include the independently guarded bool, String, or block-quant paths, and no out-of-bounds write or code execution was demonstrated. This issue is fixed in versions 0.21.16, 0.22.2, and 0.23.1.16hCVE-2026-72085.3 MED—
———Yealink SIP-T33G firmware versions 124.86.x.x prior to 124.87.0.0 contain a race condition vulnerability that allows authenticated attackers to interrupt active diagnostic processes by concurrently deleting output files written to predictable paths under the diagnostic directory. Attackers can trigger a diagnostic operation such as traceroute or ping and simultaneously invoke the file deletion endpoint to terminate the running process, leaving the system in an inconsistent state.17hCVE-2026-552445.0 MED—
———ASTEVAL is an evaluator of Python expressions and statements. Prior to 1.0.9, FROM_PY in asteval/astutils.py exposes BaseException, SystemExit, KeyboardInterrupt, and GeneratorExit to expressions evaluated by asteval.Interpreter.eval(), while run() and eval() in asteval/asteval.py catch Exception rather than these non-Exception BaseException subclasses. When an attacker-controlled expression raises one of these classes, on_raise() passes the class to raise_exception(), and the resulting exception bypasses the interpreter's safety handlers and propagates into the calling application. A consuming service that evaluates untrusted expressions can therefore be terminated or have signal and cleanup handling disrupted, causing denial of service. The separately documented read-only open() capability is not part of this vulnerability. This issue is fixed in version 1.0.9.16hCVE-2026-734967.7 HIG—
———MCP Atlassian is a Model Context Protocol (MCP) server for Atlassian products (Confluence and Jira). Prior to 0.22.0, the confluence_upload_attachment and confluence_upload_attachments tools pass a client-controlled file_path through src/mcp_atlassian/confluence/attachments.py upload_attachment, and the jira_update_issue attachments parameter reaches src/mcp_atlassian/jira/attachments.py upload_attachment, without confining either path to an approved server workspace. In a remote HTTP, SSE, or multi-user deployment, absolute or traversing paths are resolved on the MCP server and uploaded to Atlassian, allowing a client with write-tool access to disclose server files, environment-held Atlassian credentials, or another tenant's data. A local single-user stdio deployment does not cross this trust boundary because the server runs in the caller's environment. This issue is fixed in version 0.22.0.16h