CVE-20XX-?
agent?

Just published.

The newest CVEs, straight from CVE.org and NVD.

16,497 CVEs · updated 2026-11-21 05:29 UTC

Reset all
  1. CVE-2026-123834

    In JetBrains Datalore before 2026.4 improper scope validation in database tasks allowed using another workspace's service account

    Auth & access Server Jetbrains Datalore

    7.5 HIGH 0.3% exploit odds Fix listed
  2. CVE-2026-123835

    In JetBrains Datalore before 2026.4 code execution was possible in another user's computation via an init script upload race condition

    Code execution Library Jetbrains Datalore

    5.9 MEDIUM 0.7% exploit odds Fix listed
  3. CVE-2026-123836

    In JetBrains Datalore before 2026.4 workspace collaborators could obtain the owner's SSH private key via Git backup

    Other Library Jetbrains Datalore

    8.8 HIGH 0.5% exploit odds Fix listed
  4. CVE-2026-122053

    The Symfony Serializer component ObjectNormalizer treats any attribute name containing . (and, from Symfony 6.4, [) as a writable property path, allowing untrusted input to modify properties of nested objects and bypass their #[Ignore] metadata attribute in Symfony versions 5.4.41 prior to 5.4.56, 6.4.9 prior to 6.4.50, 7.0.9 prior to 7.1.0, 7.1.2 prior to 7.4.23, and 8.x prior to 8.1.11. ObjectNormalizer::isAllowedAttribute() accepts such names, and setAttributeValue() hands the name straight to PropertyAccessor::setValue(), which parses it as a property path and traverses into the nested object. A payload key such as customer.admin writes to $order->customer->admin. The #[Ignore] metadata attribute is only consulted when the nested object is denormalized in its own right, never when it is reached through a path. #[Groups] restrictions are not bypassed: the path is only accepted when no groups are passed in the context, and then groups restrict nothing. When the root class carries neither #[Ignore] nor #[Groups], no allowlist applies and every key of the payload is let through. Because PropertyAccessor also resolves a path segment against public methods, a segment matching a public method name invokes that method with the attacker’s value.Applications that bind untrusted input to objects are affected, including controllers using #[MapRequestPayload] or #[MapQueryString], whose default normalizer is ObjectNormalizer. With OBJECT_TO_POPULATE, as used for partial updates, the traversal reaches an already hydrated entity graph, and the modified field may be persisted. The fix accepts a property path only when the class metadata declares an attribute of that exact name. Declared paths keep working unchanged, and a path in the payload with no matching declaration is ignored.

    Other Web app

    6.9 MEDIUM 0.3% exploit odds Fix listed
  5. CVE-2026-122193

    OS command injection in the example SSM document provisioning script in Trusted Remote Execution (REX) before 2026-07-28 might allow an authenticated remote user who is permitted to invoke the Rex-RO or Rex-RW SSM document to execute arbitrary commands as root on the managed instance, bypassing Cedar policy enforcement. To remediate this issue, users should delete any Rex-RO and Rex-RW SSM documents created from the example script, in each account and Region where it was run.

    Injection

    9.3 CRITICAL 2% exploit odds Fix listed
  6. CVE-2026-118515

    The Symfony Serializer component DataUriNormalizer handles only the exact types SplFileInfo, SplFileObject, and File, so for a property typed with a subclass such as UploadedFile, ObjectNormalizer instantiates the object through its constructor from attacker-controlled data, including the file path. This allows an arbitrary file on the server to be treated as an uploaded file in Symfony versions prior to 5.4.56, 6.x prior to 6.4.50, 7.x prior to 7.4.23, and 8.x prior to 8.1.11. A request body such as {"image": {"path": "/path/to/project/.env.local", "originalName": "a.png", "test": true}} makes the application handle any file on the server as an upload, without the file being sent. With test: true, UploadedFile::isValid() skips is_uploaded_file() and move() falls back to rename(). Assert\File does not prevent this because it validates the file at the attacker-chosen path, which exists on the server. Any code that deserializes untrusted data into a type that holds such a property is affected. That includes #[MapRequestPayload] on a DTO with an UploadedFile property, which is how uploads are mapped in Symfony 8.1; a single unauthenticated request with a JSON or form body is enough. Depending on what the application does with the upload, this can expose any file readable by PHP (for instance, when uploads are stored and served back), move any such file when move() is called, and reveal whether a path exists. A serializer built without DataUriNormalizer ahead of the object normalizer is not protected by the fix; the serializer configured by FrameworkBundle is.

    Disclosure Web app

    8.4 HIGH 0.6% exploit odds Fix listed
  7. CVE-2026-106136

    Cross-Site Scripting (XSS) in the Chat component (@progress/kendo-react-conversational-ui) in Progress Software KendoReact versions 0.6.0 through 16.1.0 allows a low-privileged attacker who can supply Chat messages to execute arbitrary JavaScript in the origin of the application via an unsafe URL in an openUrl quick action, which runs when a user clicks the action.

    XSS Web app

    5.4 MEDIUM 0.2% exploit odds No fix listed
  8. CVE-2026-106141

    Cross-Site Scripting (XSS) in the Editor component (@progress/kendo-angular-editor) in Progress Software Kendo UI for Angular versions prior to 25.3.0 allows a low-privileged attacker who can supply Editor content to execute arbitrary JavaScript in another user's browser via a link or image with an unsafe URL in the Editor content, which runs when the user interacts with it.

    XSS Web app

    5.4 MEDIUM 0.2% exploit odds No fix listed
  9. CVE-2026-106142

    Cross-Site Scripting (XSS) in the Editor component (@progress/kendo-react-editor) in Progress Software KendoReact versions 2.7.0 through 16.1.0 allows a low-privileged attacker who can supply Editor content to execute arbitrary JavaScript in another user's browser via a link with an unsafe URL in the Editor content, which runs when the user clicks the link.

    XSS Web app

    5.4 MEDIUM 0.2% exploit odds No fix listed
  10. CVE-2026-106143

    Cross-Site Scripting (XSS) in the Editor component (@progress/kendo-vue-editor) in Progress Software Kendo UI for Vue versions 2.7.1 through 10.0.0 allows a low-privileged attacker who can supply Editor content to execute arbitrary JavaScript in another user's browser via an unsafe URL entered in the Insert Link or Insert Image dialog, which runs when the user clicks the resulting link.

    XSS Web app

    5.4 MEDIUM 0.1% exploit odds No fix listed
  11. CVE-2026-106144

    Prototype Pollution in the Scheduler component (@progress/kendo-angular-scheduler and @progress/kendo-angular-common) in Progress Software Kendo UI for Angular versions prior to 25.1.0 allows a high-privileged attacker who can control a field path processed by the Scheduler to modify the prototype of built-in objects, resulting in a limited impact on confidentiality, integrity, and availability.

    Other Web app

    4.1 MEDIUM 0.2% exploit odds No fix listed
  12. CVE-2026-104656

    Improper neutralization of input in the sessions grid in Progress Telerik Fiddler Everywhere 8.3.0 and earlier allows an attacker to execute backend IPC commands by convincing a user to open a specially crafted SAZ file.

    XSS

    7.3 HIGH 0.2% exploit odds No fix listed
  13. CVE-2026-104657

    Missing integrity verification in the Progress Software Fiddler Download and Setup Skill on Linux, macOS, and Windows which allows an attacker to compromise or influence a release component and execute attacker-controlled code when a user runs the skill. On macOS, the skill stages the package at a predictable path and installs it with administrator privileges, which may allow code execution with root privileges after user approval. A local process may also replace the staged macOS package while the skill waits for approval.

    Other OS & kernel

    7.5 HIGH 0.2% exploit odds No fix listed
  14. CVE-2026-106134

    Cross-Site Scripting (XSS) in the Menu component (@progress/kendo-react-layout) in Progress Software KendoReact versions 1.3.0 through 16.1.0 allows a low-privileged attacker who can influence the Menu item data to execute arbitrary JavaScript in the origin of the application via an unsafe URL in a Menu item, which runs when a user selects the item.

    XSS Web app

    5.4 MEDIUM 0.2% exploit odds No fix listed
  15. CVE-2026-106135

    Cross-Site Scripting (XSS) in the Chat component (@progress/kendo-angular-conversational-ui) in Progress Software Kendo UI for Angular versions prior to 25.3.0 allows a low-privileged attacker who can supply Chat messages to execute arbitrary JavaScript in another user's browser via an unsafe URL in an openUrl quick action, which runs when the user clicks the action.

    XSS Web app

    5.4 MEDIUM 0.2% exploit odds No fix listed
  16. CVE-2026-106127

    Prototype Pollution in the setter utility (@progress/kendo-react-common) in Progress Software KendoReact versions 3.5.0 through 16.0.0 allows a high-privileged attacker who can control a field path passed to the utility, directly or through components such as the Form, Scheduler, and Gantt, to modify the prototype of built-in objects, resulting in a limited impact on confidentiality, integrity, and availability.

    Other Web app

    4.1 MEDIUM 0.2% exploit odds No fix listed
  17. CVE-2026-106128

    Prototype Pollution in @progress/kendo-vue-common in Progress Software Kendo UI for Vue versions 0.4.0 through 9.2.0 allows a high-privileged attacker who can control a field path processed by components such as the Scheduler to modify the prototype of built-in objects, resulting in a limited impact on confidentiality, integrity, and availability.

    Other Web app

    4.1 MEDIUM 0.2% exploit odds No fix listed
  18. CVE-2026-103984

    FreeScout is a self-hosted help desk and shared mailbox. Prior to 1.8.231, when limit_user_customer_visibility is enabled, the load_customer_info AJAX action performs a global lookup using an attacker-controlled customer email and checks only whether the agent can access a separately supplied mailbox. A non-admin agent can name an accessible mailbox and retrieve an unrelated customer's rendered profile without invoking CustomerPolicy::view(). The response can expose names, email addresses, phone numbers, company and job information, postal addresses, social profiles, websites, and confidential internal agent notes. This issue is fixed in version 1.8.231.

    Auth & access Web app

    6.5 MEDIUM 0.4% exploit odds Fix listed
  19. CVE-2026-103985

    FreeScout is a self-hosted help desk and shared mailbox. Prior to 1.8.231, the profile form exposes the only_assigned_tickets permission to its non-admin owner, and UsersController::profileSave() adds or removes User::PERM_ONLY_ASSIGNED_TICKETS without checking that the editor is an administrator. A restricted agent can uncheck the option, or submit an ordinary profile save that omits it, to remove the administrator-imposed restriction. The agent can then read and modify conversations assigned to other agents in every mailbox they already belong to. This issue is fixed in version 1.8.231.

    Auth & access

    7.1 HIGH 0.3% exploit odds Fix listed
  20. CVE-2026-90436

    ** UNSUPPORTED WHEN ASSIGNED ** Deserialization of an inbound SOAP fault in Apache Axis 1.x allows unauthenticated remote class instantiation. When a SOAP <Fault> carries a <detail><exceptionName> element, Axis loads the named class and, if a String <data> element is also present, instantiates it through its String constructor before checking that it is an AxisFault. Loading alone runs the class's static initializer. With a suitable class on the application's classpath this leads to remote code execution. The fault is parsed while the engine resolves the target service, before any per-service handler chain runs, so authentication handlers configured on a service do not protect against it. This affects all Apache Axis 1.x versions through 1.4. As Axis 1 has been EOL we recommend you migrate to a different SOAP engine, such as Apache Axis 2/Java. Alternatively you could use a build of Axis with the patch from https://github.com/apache/axis-axis1-java/commit/41f3b4e8d04e310ca81f86b92630ab2c11d608bd applied. The Apache Axis project does not expect to create an Axis 1.x release fixing this problem.

    exploited Code execution Web app Apache Axis

    9.8 CRITICAL 2% exploit odds Fix listed
  21. CVE-2026-80302

    Insufficient isolation in the Avast sandbox file virtualization layer on Windows allows a local, low-privileged attacker executing inside the sandbox to write arbitrary files outside sandbox containment and execute code on the host. The sandbox virtualizes file system operations by intercepting write requests to protected system directories. However, the virtualization implementation fails to properly validate write targets in certain edge cases, allowing writes to bypass the virtualization redirection and reach the actual host file system in protected locations. An attacker can exploit this to overwrite executable files, system libraries, or startup configurations with arbitrary content, leading to code execution with host privileges on subsequent system operations or reboot.

    Code execution Browser & mail

    8.8 HIGH 0.1% exploit odds No fix listed
  22. CVE-2026-80303

    An integer underflow vulnerability in the Avast Windows Filtering Platform (WFP) stream parser on Windows allows denial of service by causing a system crash. The WFP callout driver processes network stream data and computes length values for packet fragment reassembly. The length calculation does not properly validate input stream sizes against expected boundaries. When processing specially crafted input, the length computation underflows, resulting in allocation or buffer access with an extremely large wrapped-around size value. The underflow can be triggered remotely via network traffic destined for the filtered port, or locally via crafted data provided to the filtering interface. The resulting memory allocation failure or out-of-bounds access causes a kernel panic and system crash.

    Denial of service OS & kernel

    6.2 MEDIUM 0.1% exploit odds No fix listed
  23. CVE-2026-66987

    HCL AION is affected by a vulnerability where the account lockout mechanism did not sufficiently restrict repeated failed login attempts, allowing an attacker to make a large number of password guesses against a user account. This may increase the risk of account compromise through brute-force attacks, potentially resulting in unintended behavior or security impact under certain conditions.

    Other

    5.3 MEDIUM 0.2% exploit odds No fix listed
  24. CVE-2026-66986

    HCL AION is affected by a vulnerability where user sessions did not expire after a sufficient period of inactivity. An unattended authenticated session could be used by another person with access to the same system to view information or perform actions on behalf of the user, potentially resulting in unintended behavior or security impact under certain conditions.

    Other

    3.3 LOW 0.1% exploit odds No fix listed
  25. CVE-2026-80301

    Improper isolation in the Avast sandbox on Windows allows a local, low-privileged attacker executing inside the sandbox to access COM interfaces to the Task Scheduler service, escape containment, and execute code with elevated privileges outside the sandbox. The sandbox virtualizes file system and registry access but does not adequately restrict COM interface access to Task Scheduler. A process inside the sandbox can instantiate COM objects bound to Task Scheduler functionality and invoke methods to schedule and execute tasks on the host system. The Task Scheduler COM pathway is not blocked by the sandbox policy, allowing arbitrary code execution in the host security context. This escape vector bypasses the sandbox's intended process and file isolation boundaries and can be chained with other host-level privilege escalation techniques.

    Code execution

    8.8 HIGH 0.1% exploit odds No fix listed
  26. CVE-2026-66982

    HCL AION is affected by a vulnerability where browser autocomplete was enabled on sensitive login form fields, allowing the browser to store and suggest previously entered values. A malicious user with access to the same browser or system could retrieve saved credentials, potentially resulting in unintended behavior or security impact under certain conditions.

    Other Browser & mail

    3.3 LOW 0.1% exploit odds No fix listed
  27. CVE-2026-66981

    HCL AION is affected by a vulnerability where the application trusted the value of the HTTP Host header supplied in requests without validation. An attacker could manipulate this header to influence links, redirects or responses generated by the application, potentially resulting in unintended behavior or security impact under certain conditions.

    Other Web app

    6.5 MEDIUM 0.1% exploit odds No fix listed
  28. CVE-2026-66984

    HCL AION is affected by a vulnerability where certain application responses revealed internal server file system paths. This information could help an attacker understand the application structure and plan further attacks, potentially resulting in unintended behavior or security impact under certain conditions.

    Disclosure Server

    3.1 LOW 0.2% exploit odds No fix listed
  29. CVE-2026-58141

    Dify before 1.15.0 contains a server-side request forgery vulnerability that allows any authenticated console user to bypass SSRF protections and issue arbitrary HTTP requests to internal services and cloud metadata endpoints due to the Squid proxy configuration only defining source ACLs without destination deny rules and the make_request() helper performing no independent destination IP validation. Attackers can use endpoints such as GET /console/api/remote-files/<url>, POST /console/api/remote-files/upload, the HTTP Request workflow node, website crawl, or document import from URL to reach internal components like the sandbox, internal APIs, and PostgreSQL, and on AWS deployments read IAM credentials from the instance metadata service, with the HTTP Request node further enabling POST, PUT, and DELETE writes to internal services relying on network isolation for authorization.

    Disclosure Server

    7.2 HIGH 0.4% exploit odds Fix listed
  30. CVE-2026-21834

    HCL AION is affected by a vulnerability where the Content-Security-Policy (CSP) security response header was missing from application responses. Without this header, the browser does not restrict the sources of scripts and other content, increasing exposure to cross-site scripting and content-injection attacks, potentially resulting in unintended behavior or security impact under certain conditions.

    XSS Web app

    3.1 LOW 0.2% exploit odds No fix listed
  31. CVE-2026-123792

    The Symfony Yaml components Inline and Parser, parse some YAML documents inefficiently, so that a small crafted document can take disproportionately long to process, keeping a PHP process busy for seconds or minutes, potentially causing a denial of service in Symfony versions prior to 5.4.56, 6.x prior to 6.4.50, 7.x prior to 7.4.23, and 8.x prior to 8.1.11. The parser repeats work it has already done in several places. An embedded mapping in a flow sequence is parsed again for every scalar appended to it, and so are the collections nested in it, so the parse time grows exponentially with the nesting depth. The patterns matching mapping keys and tags scan a run of spaces once per character of the run, and each scalar of a flow collection is matched against a copy of the rest of the collection, so the parse time grows quadratically with the size of the input. Nested blocks are also copied at each nesting level, which consumes memory as well as time. A few other constructs, such as an empty sequence item followed by a blank line or an anchor named 0, trigger similar repeated work. Any code parsing untrusted YAML is affected, with the default flags. In Symfony 7.4 and later, that includes #[MapRequestPayload] when the Yaml component is installed and the endpoint does not restrict acceptFormat. A single unauthenticated request with a YAML content type keeps a PHP worker busy until max_execution_time. JSON documents are valid YAML, so a JSON payload sent with a YAML content type is enough for the flow collection case. In Symfony 6.4, #[MapRequestPayload] accepts YAML only when the application registers a yaml request format.

    Denial of service Web app

    8.7 HIGH 0.5% exploit odds Fix listed
  32. CVE-2026-123677

    A vulnerability was found in xxyopen/201206030 novel up to 3.5.0. Affected is the function getBookChapter of the file src/main/java/io/github/xxyopen/novel/service/impl/BookServiceImpl.java of the component Book Chapter Read Handler. Performing a manipulation of the argument chapterId results in authorization bypass. The attack may be initiated remotely. The exploit has been made public and could be used. The patch is named b8134e774820e0971fa9d8074405b49a01c99bdc. It is suggested to install a patch to address this issue.

    Auth & access Web app

    2.1 LOW 0.3% exploit odds Fix listed
  33. CVE-2026-12067

    Missing exception handlers in AvastSvc native message dispatcher (ashServ.dll) on Windows allow a local, unprivileged attacker to crash the service by sending specially crafted JSON messages. The vulnerability spans multiple handler pathways in the RPC-based native message protocol: * Log_NmSvc.message handler: converts JSON fields to strings without type validation and without exception boundaries.  * OpenInSz_NmSvc handler: converts URL and origin fields to strings without type checking, exposing an exception. * OpenInSz_NmSvc.profile handler: converts the profile field to an integer without validation. In all three cases, the exceptions are not caught by action-specific handlers and propagate up the dispatcher. The common root cause is the absence of try-catch blocks around JSON type conversions in the native message processing chain.

    Denial of service

    6.5 MEDIUM 0.3% exploit odds No fix listed
  34. CVE-2026-123623

    Grida through 0.0.24 contains a missing authorization vulnerability in the Forms file-preview API routes that allows unauthenticated attackers to read other organizations' form response files. Attackers can use form and field IDs exposed in public form payloads with a chosen path parameter to obtain service-role signed URLs for attachments and connected Supabase bucket objects.

    Auth & access Web app

    8.2 HIGH 0.4% exploit odds Fix listed
  35. CVE-2026-123675

    A flaw has been found in OpenBMB Internet of Agents up to 9d78e88e47160eb42ee1ee8538c05e1c2c47877a. This affects the function execute_code of the file im_client/agents/tools/code_executor.py of the component Code Executor Tool. This manipulation of the argument filename/lang causes code injection. The attack can be initiated remotely. The exploit has been published and may be used. Continious delivery with rolling releases is used by this product. Therefore, no version details of affected nor updated releases are available. The project was informed of the problem early through an issue report but has not responded yet.

    Injection

    5.5 MEDIUM 0.5% exploit odds No fix listed
  36. CVE-2026-123676

    A vulnerability has been found in xxyopen/201206030 novel up to 3.5.0. This impacts the function BookServiceImpl.getBookContentAbout of the file src/main/java/io/github/xxyopen/novel/service/impl/BookServiceImpl.java of the component Book Content Service. Such manipulation of the argument chapterId leads to improper authorization. The attack can be launched remotely. The exploit has been disclosed to the public and may be used. The name of the patch is 966c4bdde3521dc0a165d15a2acff66ee33eaaa7. Applying a patch is advised to resolve this issue.

    Auth & access Web app

    5.5 MEDIUM 0.3% exploit odds Fix listed
  37. CVE-2026-121881

    SQL injection vulnerability in the get_last_seen code-server session-state RPC in Posit Software, PBC Posit Workbench before 2026.08.0 allows authenticated users to read arbitrary data from the Workbench database via the sessionIds parameter. NOTE: exploitation requires a valid authenticated session; no administrative privileges are needed. The vulnerability is limited to disclosure of database contents.

    Injection Server

    5.7 MEDIUM 0.3% exploit odds No fix listed
  38. CVE-2026-121880

    Authorization bypass through user-controlled key in the localhost port-proxy (/p/<token>/) in Posit Software, PBC RStudio Server and Posit Workbench before 2026.08.0 allows remote authenticated users to reach web applications and other services bound to 127.0.0.1 under a different user's session, and to view their live rendered content and data, via a proxy URL constructed from the attacker's own port token for an arbitrary port number. The port token is a documented obfuscation measure and not an authorization boundary, and the proxy does not verify that the requested port belongs to the requesting user's session. NOTE: exploitation requires an authenticated account on the same multi-user server and a target service listening on localhost at the time of access. Impact may be greater on hosts that co-locate additional privileged localhost services, and the bypass is not prevented by allow-shell=false; no setting restricts the port-proxy without disabling all application proxying.

    Auth & access Server

    4.7 MEDIUM 0.3% exploit odds No fix listed
  39. CVE-2026-121623

    The Symfony Serializer component DataUriNormalizer::denormalize() matches the data: scheme with a case-sensitive regular expression, allowing a value with an uppercase scheme, such as DATA:,/../../../etc/passwd, to bypass it, resulting in Arbitrary Local File Read through Path Traversal in Symfony versions prior to 5.4.56, 6.x prior to 6.4.50, 7.x prior to 7.4.23, and 8.x prior to 8.1.11. A value whose scheme is not lowercase data: are treated as a local file path and opened instead of being decoded as inline data, so the resulting File, \SplFileInfo, or \SplFileObject points at a file on the server, which the application then handles as the uploaded content. Any endpoint that deserializes untrusted input into one of these types is affected. FrameworkBundle registers DataUriNormalizer ahead of ObjectNormalizer by default. A single request with a crafted DATA: value is enough to read arbitrary local files that the application can access, with no authentication needed if the endpoint is public.

    Disclosure Web app

    8.2 HIGH 0.6% exploit odds Fix listed
  40. CVE-2026-111573

    Incorrect permissions on the Checkmk Windows agent's data directory, caused by a race condition that discards the permission changes during service start, in Checkmk <2.5.0p18, <2.4.0p39, <2.3.0p52 and 2.2.0 (EOL) allow a local user to execute arbitrary code as SYSTEM by placing scripts in the agent's plugins or local folders, and to read the agent controller's private key.

    Code execution OS & kernel

    6.3 MEDIUM 0.1% exploit odds No fix listed
  41. CVE-2026-87045

    The HTTP/2 server in subsys/net/lib/http/http_server_http2.c clears the connection's stream binding at the start of every frame. handle_http_frame_header() executed client->current_stream = NULL unconditionally, but a CONTINUATION frame does not open a new stream and enter_http_frame_continuation_state() never re-bound one, so the rest of the header block was processed with client->current_stream unset. Two paths then write through the NULL pointer. A request for an unregistered resource, or a method the matched resource does not support, is answered by send_http2_404() or send_http2_405(), and both call send_headers_frame(), which unconditionally sets client->current_stream->headers_sent = true. A POST, PUT or PATCH to a registered dynamic resource reaches handle_http2_dynamic_resource(), which unlike its siblings (handle_http2_static_resource(), dynamic_get_del_opts_req_v2(), dynamic_post_put_req_v2()) had no client->current_stream == NULL guard, and stores through the pointer with client->current_stream->current_detail = dynamic_detail. Any peer that can reach the server can trigger this without authentication or prior state. A single stray CONTINUATION frame with END_HEADERS set, sent right after the client preface and requesting any unregistered path, is enough, because nothing rejected a CONTINUATION that no HEADERS frame had opened. A legitimate split header block (a HEADERS frame with END_HEADERS clear whose payload ends on a header-field boundary, then a CONTINUATION with END_HEADERS set) reaches the same state. The in-tree test test_http2_post_headers_with_continuation() did not catch this because it splits the header block mid field, which takes the handle_incomplete_http_header() shortcut that keeps the stream bound. The result is a write to a fixed low address (a small offset into struct http2_stream_ctx), which faults on mainstream targets (flash/vector-table region on Cortex-M, or a trap under MPU/MMU null-pointer detection), and the default fatal-error handler halts the system. Impact is an unauthenticated, remotely triggerable denial of service against any device running the HTTP/2 server: the 404 path needs no particular resource to be registered, and HTTP/2 is built into CONFIG_HTTP_SERVER with cleartext h2c accepted, so TLS does not prevent it. Neither the target address nor the written value is attacker-controlled, so there is no confidentiality or integrity impact. Applications built without CONFIG_HTTP_SERVER are not affected.

    Memory safety Server

    7.5 HIGH 0.4% exploit odds Fix listed
  42. CVE-2026-59105

    Jumis (formerly Tildes Jumis) creates and uses a Microsoft SQL Server login named ##Tilde_JumisAdmin## whose password is derived deterministically from a single non-secret input - the SQL Server name (@@servername). The derivation algorithm is hard-coded in the client application and is identical across every installation; it contains a hard-coded secret key for entropy. Because the only input (the server name) is not secret - the instance suffix is hard-coded and the host name is discoverable via the SQL Server Browser service (UDP 1434), the TDS pre-login response, or DNS/NetBIOS - any party able to reach a Jumis SQL Server can compute the password offline and authenticate. The ##Tilde_JumisAdmin## login is a member of the sysadmin server role. Successful authentication therefore grants full administrative control of the SQL Server instance and read/write access to every customer's accounting database hosted on it. The on-premise server installation also programmatically enables xp_cmdshell during routine operation, so sysadmin access escalates to operating-system command execution on the database server.

    Auth & access Server

    9.3 CRITICAL 1% exploit odds No fix listed
  43. CVE-2026-64894

    Inclusion of functionality from untrusted control sphere vulnerability in Johnson Controls EasyIO Neo IOManager and Tools Installer allows Malicious Software Update. This issue affects EasyIO Neo IOManager and Tools Installer: 1.1.0.0.

    Other Device

    7.3 HIGH 0.1% exploit odds No fix listed
  44. CVE-2026-53980

    Charmbracelet wish before 2.0.5 contains a path traversal vulnerability in the git middleware (charm.land/wish/v2) that allows authenticated SSH users to escape the configured repository directory by supplying crafted repo paths that bypass a flawed forward-slash-count containment check in git.Middleware(). Attackers can use a single-depth path ../evil, or on Windows a payload ..\..\evil that defeats the slash-count guard entirely, to read or write arbitrary repositories outside the intended root via git-upload-pack, git-upload-archive, or git-receive-pack.

    Disclosure OS & kernel

    8.6 HIGH 0.4% exploit odds Fix listed
  45. CVE-2026-123783

    Skyvern through 1.0.56 contains a sandbox bypass vulnerability that allows authenticated users to execute code by reconstructing functions from raw bytecode with types.CodeType. Attackers can deploy scripts that evade the is_safe_script_code AST denylist and import blocked modules such as subprocess or socket to run code in the Skyvern server process.

    Code execution Server

    8.7 HIGH 0.6% exploit odds Fix listed
  46. CVE-2026-123784

    Fluxbox through 1.3.7 contains an OS command injection vulnerability in RootTheme::reconfigTheme() that allows attackers to execute commands via crafted style background values. Attackers can embed shell metacharacters in background pixmap paths, random directories, or gradient texture text, which are concatenated unescaped into an fbsetbg command run via $SHELL -c when the style loads.

    Injection

    8.5 HIGH 0.7% exploit odds Fix listed
  47. CVE-2026-123785

    Fluxbox through 1.3.7 contains an infinite loop vulnerability in Menu::updateMenu() in src/FbTk/Menu.cc that allows attackers to hang the window manager via crafted style geometry values. Attackers can supply a style with an oversized unbounded menu.titleHeight so the column loop never terminates, causing sustained CPU use and denial of service at startup.

    Denial of service

    6.8 MEDIUM 0.2% exploit odds Fix listed
  48. CVE-2026-123786

    Agenta before 0.122.5 contains a server-side request forgery vulnerability that allows workspace members with workflow edit rights to redirect workflow invocations by setting arbitrary revision URLs. Attackers can capture the attached secret-resolve Authorization token to read provider secrets from the vault API, or reach internal and cloud metadata addresses.

    Disclosure Server

    8.4 HIGH 0.2% exploit odds Fix listed
  49. CVE-2026-123664

    A security flaw has been discovered in nasa-gcn gcn.nasa.gov up to b299b562af665878964b981a459f0a9c04b11adf. This issue affects some unknown processing of the file /user/email of the component Email Notification Handler. Performing a manipulation of the argument Recipient results in improper authorization. The attack is possible to be carried out remotely. The exploit has been released to the public and may be used for attacks. This product adopts a rolling release strategy to maintain continuous delivery. Therefore, version details for affected or updated releases cannot be specified. The pull request to fix this issue awaits acceptance.

    Auth & access

    2.1 LOW 0.2% exploit odds Fix listed
  50. CVE-2026-123665

    A weakness has been identified in kubescape up to 4.0.15. Impacted is the function GetWriter of the file core/pkg/resultshandling/printer/printresults.go of the component Output. Executing a manipulation of the argument outputFile can lead to permission issues. The attack requires local access. Upgrading to version 4.0.16 is recommended to address this issue. This patch is called d90f77314c7d4f2b53cf2425844bde2315bd26d2. Upgrading the affected component is advised.

    Disclosure Device

    2.4 LOW 0.1% exploit odds Fix listed

The newest published CVE records from CVE.org and NVD, refreshed every few minutes, with severity, exploit odds and fix status.