Overview 16,255 matches · 202 in Waratek scope · 33 protected · 197 exploited · 46 CISA KEV · 295 public PoC · 35 EPSS ≥ 0.5
16,255matches
202in Waratek scope1.2%
33protected0.2%
197exploited1.2%
46CISA KEV0.3%
295public PoC1.8%
35EPSS ≥ 0.50.2%
Protected by Waratek
33
0.2%
Rule in development
0
0.0%
Mitigation candidate
9
0.1%
No exploit published
160
1.0%
Outside Waratek scope
62
0.4%
Not applicable
14,888
91.6%
Withdrawn CVE
1,103
6.8%
Critical
1,663
10.2%
High
5,966
36.7%
Medium
6,183
38.0%
Low
1,211
7.5%
None
33
0.2%
Unknown
1,199
7.4%
blocked by ARMR today
33
0.2%
partly blocked
0
0.0%
not blocked
16,221
99.8%
not established
1
0.0%
unrecorded
0
0.0%
split by
peak 6,304 / month
Jan 265,143
Protected by Waratek14
Mitigation candidate1
No exploit published64
Outside Waratek scope12
Not applicable4,211
Withdrawn CVE841
Critical487
High1,661
Medium1,820
Low273
Unknown902
blocked by ARMR today14
not blocked5,129
Exploited75
Working exploit published33
Proof of concept only1,120
Forecast only1
No public exploit3,914
Feb 264,808
Protected by Waratek7
Mitigation candidate2
No exploit published33
Outside Waratek scope6
Not applicable4,568
Withdrawn CVE192
Critical470
High1,668
Medium1,973
Low439
None31
Unknown227
blocked by ARMR today7
not blocked4,801
Exploited58
Working exploit published47
Proof of concept only1,264
Forecast only2
No public exploit3,437
Mar 266,304
Protected by Waratek12
Mitigation candidate6
No exploit published63
Outside Waratek scope44
Not applicable6,109
Withdrawn CVE70
Critical706
High2,637
Medium2,390
Low499
None2
Unknown70
blocked by ARMR today12
not blocked6,291
not established1
Exploited64
Working exploit published54
Proof of concept only1,830
No public exploit4,356
Jan 26Feb 26Mar 26
| CVE | Severity | Published▾ | Status | Exploitation | Description |
|---|---|---|---|---|---|
| CVE-2026-5237 | Medium | 2026-03-31 | Not applicable | No public exploit | A security flaw has been discovered in itsourcecode Payroll Management System 1.0. Affected by this vulnerability is an unknown functionality of the file /manage_user.php of the component Parameter Handler. Performing a manipulation of the argument ID results in sql injection. The attack is possible to be carried out remotely. The exploit has been released to the public and may be used for attacks. |
| CVE-2026-5236 | Low | 2026-03-31 | Not applicable | Proof of concept only | A vulnerability was identified in Axiomatic Bento4 up to 1.6.0-641. Affected is the function AP4_BitReader::SkipBits of the file Ap4Dac4Atom.cpp of the component DSI v1 Parser. Such manipulation of the argument n_presentations leads to heap-based buffer overflow. The attack needs to be performed locally. The exploit is publicly available and might be used. The project was informed of the problem early through an issue report but has not responded yet. |
| CVE-2026-5235 | Low | 2026-03-31 | Not applicable | Proof of concept only | A vulnerability was determined in Axiomatic Bento4 up to 1.6.0-641. This impacts the function AP4_BitReader::ReadCache of the file Ap4Dac4Atom.cpp of the component MP4 File Parser. This manipulation causes heap-based buffer overflow. The attack needs to be launched locally. The exploit has been publicly disclosed and may be utilized. The project was informed of the problem early through an issue report but has not responded yet. |
| CVE-2026-5215 | Low | 2026-03-31 | Not applicable | Proof of concept only | A vulnerability was identified in D-Link DNS-120, DNR-202L, DNS-315L, DNS-320, DNS-320L, DNS-320LW, DNS-321, DNR-322L, DNS-323, DNS-325, DNS-326, DNS-327L, DNR-326, DNS-340L, DNS-343, DNS-345, DNS-726-4, DNS-1100-4, DNS-1200-05 and DNS-1550-04 up to 20260205. The impacted element is the function cgi_get_ipv6 of the file /cgi-bin/network_mgr.cgi. Such manipulation leads to improper access controls. The exploit is publicly available and might be used. |
| CVE-2026-5214 | High | 2026-03-31 | Not applicable | Proof of concept only | A vulnerability was found in D-Link DNS-120, DNR-202L, DNS-315L, DNS-320, DNS-320L, DNS-320LW, DNS-321, DNR-322L, DNS-323, DNS-325, DNS-326, DNS-327L, DNR-326, DNS-340L, DNS-343, DNS-345, DNS-726-4, DNS-1100-4, DNS-1200-05 and DNS-1550-04 up to 20260205. Impacted is the function cgi_addgroup_get_group_quota_minsize of the file /cgi-bin/account_mgr.cgi. The manipulation of the argument Name results in stack-based buffer overflow. The attack may be performed from remote. The exploit has been made public and could be used. |
| CVE-2026-5213 | High | 2026-03-31 | Not applicable | Proof of concept only | A vulnerability was determined in D-Link DNS-120, DNR-202L, DNS-315L, DNS-320, DNS-320L, DNS-320LW, DNS-321, DNR-322L, DNS-323, DNS-325, DNS-326, DNS-327L, DNR-326, DNS-340L, DNS-343, DNS-345, DNS-726-4, DNS-1100-4, DNS-1200-05 and DNS-1550-04 up to 20260205. The affected element is the function cgi_adduser_to_session of the file /cgi-bin/account_mgr.cgi. This manipulation of the argument read_list causes stack-based buffer overflow. It is possible to initiate the attack remotely. The exploit has been publicly disclosed and may be utilized. |
| CVE-2026-5212 | High | 2026-03-31 | Not applicable | Proof of concept only | A vulnerability has been found in D-Link DNS-120, DNR-202L, DNS-315L, DNS-320, DNS-320L, DNS-320LW, DNS-321, DNR-322L, DNS-323, DNS-325, DNS-326, DNS-327L, DNR-326, DNS-340L, DNS-343, DNS-345, DNS-726-4, DNS-1100-4, DNS-1200-05 and DNS-1550-04 up to 20260205. This issue affects the function Webdav_Upload_File of the file /cgi-bin/webdav_mgr.cgi. The manipulation of the argument f_file leads to stack-based buffer overflow. The attack is possible to be carried out remotely. The exploit has been disclosed to the public and may be used. |
| CVE-2026-5211 | High | 2026-03-31 | Not applicable | Proof of concept only | A flaw has been found in D-Link DNS-120, DNR-202L, DNS-315L, DNS-320, DNS-320L, DNS-320LW, DNS-321, DNR-322L, DNS-323, DNS-325, DNS-326, DNS-327L, DNR-326, DNS-340L, DNS-343, DNS-345, DNS-726-4, DNS-1100-4, DNS-1200-05 and DNS-1550-04 up to 20260205. This vulnerability affects the function UPnP_AV_Server_Path_Del of the file /cgi-bin/app_mgr.cgi. Executing a manipulation of the argument f_dir can lead to stack-based buffer overflow. The attack can be executed remotely. The exploit has been published and may be used. |
| CVE-2026-5210 | Medium | 2026-03-31 | Not applicable | Proof of concept only | A vulnerability was detected in SourceCodester Leave Application System 1.0. This affects an unknown part. Performing a manipulation of the argument page results in file inclusion. Remote exploitation of the attack is possible. The exploit is now public and may be used. |
| CVE-2026-5209 | Low | 2026-03-31 | Not applicable | Proof of concept only | A security vulnerability has been detected in SourceCodester Leave Application System 1.0. Affected by this issue is some unknown functionality of the component User Management Handler. Such manipulation leads to cross site scripting. The attack may be launched remotely. The exploit has been disclosed publicly and may be used. |
| CVE-2026-5206 | Low | 2026-03-31 | Not applicable | Proof of concept only | A security vulnerability has been detected in code-projects Simple Gym Management System 1.0. This vulnerability affects unknown code of the component Payment Handler. The manipulation of the argument Payment_id/Amount/customer_id/payment_type/customer_name leads to sql injection. Remote exploitation of the attack is possible. The exploit has been disclosed publicly and may be used. |
| CVE-2026-5205 | Low | 2026-03-31 | Not applicable | No public exploit | A vulnerability was identified in chatwoot up to 4.11.2. Affected by this vulnerability is the function Webhooks::Trigger in the library lib/webhooks/trigger.rb of the component Webhook API. Such manipulation of the argument url leads to server-side request forgery. The attack can be launched remotely. The exploit is publicly available and might be used. The vendor was contacted early about this disclosure but did not respond in any way. |
| CVE-2026-5204 | High | 2026-03-31 | Not applicable | Proof of concept only | A vulnerability was determined in Tenda CH22 1.0.0.1. Affected is the function formWebTypeLibrary of the file /goform/webtypelibrary of the component Parameter Handler. This manipulation of the argument webSiteId causes stack-based buffer overflow. The attack can be initiated remotely. The exploit has been publicly disclosed and may be utilized. |
| CVE-2026-5203 | Low | 2026-03-31 | Not applicable | Proof of concept only | A vulnerability was found in CMS Made Simple up to 2.2.22. This impacts the function _copyFilesToFolder in the library modules/UserGuide/lib/class.UserGuideImporterExporter.php of the component UserGuide Module XML Import. The manipulation results in path traversal. It is possible to launch the attack remotely. The exploit has been made public and could be used. This issue has been reported early to the project. They confirmed, that "this has already been discovered and fixed for the next release." |
| CVE-2026-5201 | High | 2026-03-31 | Not applicable | Proof of concept only | A flaw was found in the gdk-pixbuf library. This heap-based buffer overflow vulnerability occurs in the JPEG image loader due to improper validation of color component counts when processing a specially crafted JPEG image. A remote attacker can exploit this flaw without user interaction, for example, via thumbnail generation. Successful exploitation leads to application crashes and denial of service (DoS) conditions. |
| CVE-2026-5198 | Medium | 2026-03-31 | Not applicable | Proof of concept only | A vulnerability was determined in code-projects Student Membership System 1.0. The impacted element is an unknown function of the file /admin/index.php of the component Admin Login. This manipulation of the argument username/password causes sql injection. Remote exploitation of the attack is possible. The exploit has been publicly disclosed and may be utilized. |
| CVE-2026-5197 | Low | 2026-03-31 | Not applicable | Proof of concept only | A vulnerability was found in code-projects Student Membership System 1.0. The affected element is an unknown function of the file /delete_user.php. The manipulation of the argument ID results in sql injection. The attack may be launched remotely. The exploit has been made public and could be used. |
| CVE-2026-5196 | Low | 2026-03-31 | Not applicable | Proof of concept only | A vulnerability has been found in code-projects Student Membership System 1.0. Impacted is an unknown function of the file /delete_member.php. The manipulation of the argument ID leads to sql injection. The attack may be initiated remotely. The exploit has been disclosed to the public and may be used. |
| CVE-2026-5195 | Medium | 2026-03-31 | Not applicable | Proof of concept only | A flaw has been found in code-projects Student Membership System 1.0. This issue affects some unknown processing of the component User Registration Handler. Executing a manipulation can lead to sql injection. The attack can be launched remotely. |
| CVE-2026-5190 | High | 2026-03-31 | Not applicable | No public exploit | Out-of-bounds write in the streaming decoder component in aws-c-event-stream before 0.6.0 might allow a third party operating a server to cause memory corruption leading to arbitrary code execution on a client application that processes crafted event-stream messages. To remediate this issue, users should upgrade to version 0.6.0 or later. |
| CVE-2026-5186 | Low | 2026-03-31 | Not applicable | No public exploit | A weakness has been identified in Nothings stb up to 2.30. This impacts the function stbi__load_gif_main of the file stb_image.h of the component Multi-frame GIF File Handler. This manipulation causes double free. The attack requires local access. The exploit has been made available to the public and could be used for attacks. The vendor was contacted early about this disclosure but did not respond in any way. |
| CVE-2026-5185 | Low | 2026-03-31 | Not applicable | No public exploit | A security flaw has been discovered in Nothings stb_image up to 2.30. This affects the function stbi__gif_load_next of the file stb_image.h of the component Multi-frame GIF File Handler. The manipulation results in heap-based buffer overflow. The attack requires a local approach. The exploit has been released to the public and may be used for attacks. The vendor was contacted early about this disclosure but did not respond in any way. |
| CVE-2026-5184 | Low | 2026-03-31 | Not applicable | Proof of concept only | A vulnerability was identified in TRENDnet TEW-713RE up to 1.02. The impacted element is an unknown function of the file /goform/setSysAdm. The manipulation of the argument admuser leads to command injection. The attack can be initiated remotely. The exploit is publicly available and might be used. The vendor was contacted early about this disclosure but did not respond in any way. |
| CVE-2026-5183 | Low | 2026-03-31 | Not applicable | Proof of concept only | A vulnerability was determined in TRENDnet TEW-713RE up to 1.02. The affected element is the function sub_421494 of the file /goform/addRouting. Executing a manipulation of the argument dest can lead to command injection. It is possible to launch the attack remotely. The exploit has been publicly disclosed and may be utilized. The vendor was contacted early about this disclosure but did not respond in any way. |
| CVE-2026-5182 | Medium | 2026-03-31 | Not applicable | Proof of concept only | A vulnerability was found in SourceCodester Teacher Record System 1.0. Impacted is an unknown function of the file Teacher Record System of the component Parameter Handler. Performing a manipulation of the argument searchteacher results in sql injection. It is possible to initiate the attack remotely. The exploit has been made public and could be used. |
| CVE-2026-5181 | Low | 2026-03-31 | Not applicable | Proof of concept only | A vulnerability has been found in SourceCodester Simple Doctors Appointment System up to 1.0. This issue affects some unknown processing of the file /doctors_appointment/admin/ajax.php?action=save_category. Such manipulation of the argument img leads to unrestricted upload. The attack may be performed from remote. The exploit has been disclosed to the public and may be used. |
| CVE-2026-5180 | Medium | 2026-03-31 | Not applicable | Proof of concept only | A flaw has been found in SourceCodester Simple Doctors Appointment System 1.0. This vulnerability affects unknown code of the file /admin/ajax.php?action=login2. This manipulation of the argument email causes sql injection. The attack is possible to be carried out remotely. The exploit has been published and may be used. |
| CVE-2026-5179 | Medium | 2026-03-31 | Not applicable | No public exploit | A vulnerability was detected in SourceCodester Simple Doctors Appointment System 1.0. This affects an unknown part of the file /admin/login.php. The manipulation of the argument Username results in sql injection. The attack can be executed remotely. The exploit is now public and may be used. |
| CVE-2026-5178 | Low | 2026-03-31 | Not applicable | Proof of concept only | A security vulnerability has been detected in Totolink A3300R 17.0.0cu.557_b20221024. Affected by this issue is the function setIptvCfg of the file /cgi-bin/cstecgi.cgi. The manipulation of the argument vlanPriLan3 leads to command injection. Remote exploitation of the attack is possible. The exploit has been disclosed publicly and may be used. |
| CVE-2026-5177 | Low | 2026-03-31 | Not applicable | Proof of concept only | A weakness has been identified in Totolink A3300R 17.0.0cu.557_b20221024. Affected by this vulnerability is the function setWiFiBasicCfg of the file /cgi-bin/cstecgi.cgi. Executing a manipulation of the argument rxRate can lead to command injection. The attack may be launched remotely. The exploit has been made available to the public and could be used for attacks. |
| CVE-2026-5176 | Medium | 2026-03-31 | Not applicable | Proof of concept only | A security flaw has been discovered in Totolink A3300R 17.0.0cu.557_b20221024. Affected is the function setSyslogCfg of the file /cgi-bin/cstecgi.cgi. Performing a manipulation of the argument provided results in command injection. The attack may be initiated remotely. The exploit has been released to the public and may be used for attacks. |
| CVE-2026-5157 | Low | 2026-03-31 | Not applicable | Proof of concept only | A vulnerability was identified in code-projects Online Food Ordering System 1.0. Affected is an unknown function of the file /form/order.php of the component Order Module. Such manipulation of the argument cust_id leads to cross site scripting. The attack may be performed from remote. The exploit is publicly available and might be used. |
| CVE-2026-5156 | High | 2026-03-31 | Not applicable | Proof of concept only | A vulnerability was determined in Tenda CH22 1.0.0.1. This impacts the function formQuickIndex of the file /goform/QuickIndex of the component Parameter Handler. This manipulation of the argument mit_linktype causes stack-based buffer overflow. The attack is possible to be carried out remotely. The exploit has been publicly disclosed and may be utilized. |
| CVE-2026-5115 | Low | 2026-03-31 | Not applicable | No public exploit | The PaperCut NG/MF (specifically, the embedded application for Konica Minolta devices) is vulnerable to session hijacking. The PaperCut NG/MF Embedded application is a software interface that runs directly on the touch screen of a multi-function device. It was internally discovered that the communication channel between the embedded application and the server was insecure, which could leak data including sensitive information that may be used to mount an attack on the device. Such an attack could potentially be used to steal data or to perform a phishing attack on the end user. |
| CVE-2026-5087 | High | 2026-03-31 | Not applicable | No public exploit | PAGI::Middleware::Session::Store::Cookie versions through 0.001003 for Perl generates random bytes insecurely. PAGI::Middleware::Session::Store::Cookie attempts to read bytes from the /dev/urandom device directly. If that fails (for example, on systems without the device, such as Windows), then it will emit a warning that recommends the user install Crypt::URandom, and then return a string of random bytes generated by the built-in rand function, which is unsuitable for cryptographic applications. This modules does not use the Crypt::URandom module, and installing it will not fix the problem. The random bytes are used for generating an initialisation vector (IV) to encrypt the cookie. A predictable IV may make it easier for malicious users to decrypt and tamper with the session data that is stored in the cookie. |
| CVE-2026-4819 | Medium | 2026-03-31 | Not applicable | No public exploit | In Search Guard FLX versions from 1.0.0 up to 4.0.1, the audit logging feature might log user credentials from users logging into Kibana. |
| CVE-2026-4818 | High | 2026-03-31 | Not applicable | No public exploit | In Search Guard FLX versions from 3.0.0 up to 4.0.1, there exists an issue which allows users without the necessary privileges to execute some management operations against data streams. |
| CVE-2026-4800 | Critical | 2026-03-31 | Not applicable | No public exploit | The fix for CVE-2021-23337 added validation for the variable option in _.template but did not apply the same validation to options.imports key names, and both paths flow into the same Function() constructor sink. When an application passes untrusted input as options.imports key names, an attacker can inject default-parameter expressions that execute arbitrary code at template compilation time. Additionally, _.template uses assignInWith to merge imports, which enumerates inherited properties via for..in, so if Object.prototype has been polluted by another vector the polluted keys are copied into the imports object and passed to Function(). The issue is fixed in 4.18.1 for lodash, lodash-es, lodash-amd, and lodash.template. As a workaround, do not pass untrusted input as key names in options.imports; use only developer-controlled, static key names. |
| CVE-2026-4799 | Medium | 2026-03-31 | Not applicable | No public exploit | In Search Guard FLX up to version 4.0.1, it is possible to use specially crafted requests to redirect the user to an untrusted URL. |
| CVE-2026-4794 | Low | 2026-03-31 | Not applicable | No public exploit | Multiple cross-site scripting (XSS) vulnerabilities in PaperCut NG/MF before 25.0.10 allow authenticated administrator users to inject arbitrary web script or HTML code via different UI fields. This could be used to compromise other admininistrator's sessions or perform unauthorized actions via the administrator's authenticated context (e.g. requires an active login session). |
| CVE-2026-4400 | High | 2026-03-31 | Not applicable | No public exploit | Insecure Direct Object Reference (IDOR) vulnerability in 1millionbot Millie chat that allows private conversations of other users being viewed by simply changing the conversation ID. The vulnerability is present in the endpoint 'api.1millionbot.com/api/public/conversations/' and, if exploited, could allow a remote attacker to access other users private chatbot conversations, revealing sensitive or confidential data without requiring credentials or impersonating users. In order for the vulnerability to be exploited, the attacker must have the user's conversation ID. |
| CVE-2026-4399 | High | 2026-03-31 | Not applicable | No public exploit | Prompt injection vulnerability in 1millionbot Millie chatbot that occurs when a user manages to evade chat restrictions using Boolean prompt injection techniques (formulating a question in such a way that, upon receiving an affirmative response ('true'), the model executes the injected instruction), causing it to return prohibited information and information outside its intended context. Successful exploitation of this vulnerability could allow a malicious remote attacker to abuse the service for purposes other than those originally intended, or even execute out-of-context tasks using 1millionbot's resources and/or OpenAI's API key. This allows the attacker to evade the containment mechanisms implemented during LLM model training and obtain responses or chat behaviors that were originally restricted. |
| CVE-2026-4317 | Critical | 2026-03-31 | Not applicable | No public exploit | SQL inyection (SQLi) vulnerability in Umami Software web application through an improperly sanitized parameter, which could allow an authenticated attacker to execute arbitrary SQL commands in the database.Specifically, they could manipulate the value of the 'timezone' request parameter by including malicious characters and SQL payload. The application would interpolate these values directly into the SQL query without first performing proper filtering or sanitization (e.g., using functions such as 'prisma.rawQuery', 'prisma.$queryRawUnsafe' or raw queries with 'ClickHouse'). The successful explotation of this vulnerability could allow an authenticated attacker to compromiso the data of the database and execute dangerous functions. |
| CVE-2026-4267 | High | 2026-03-31 | Not applicable | No public exploit | The Query Monitor – The developer tools panel for WordPress plugin for WordPress is vulnerable to Reflected Cross-Site Scripting via the ‘$_SERVER['REQUEST_URI']’ parameter in all versions up to, and including, 3.20.3 due to insufficient input sanitization and output escaping. This makes it possible for unauthenticated attackers to inject arbitrary web scripts in pages that execute if they can successfully trick a user into performing an action such as clicking on a link. |
| CVE-2026-4146 | Medium | 2026-03-31 | Not applicable | No public exploit | The Loco Translate plugin for WordPress is vulnerable to Reflected Cross-Site Scripting via the ‘update_href’ parameter in all versions up to, and including, 2.8.2 due to insufficient input sanitization and output escaping. This makes it possible for unauthenticated attackers to inject arbitrary web scripts in pages that execute if they can successfully trick a user into performing an action such as clicking on a link. |
| CVE-2026-4020 | High | 2026-03-31 | Not applicable | Exploited | The Gravity SMTP plugin for WordPress is vulnerable to Sensitive Information Exposure in all versions up to, and including, 2.1.4. This is due to a REST API endpoint registered at /wp-json/gravitysmtp/v1/tests/mock-data with a permission_callback that unconditionally returns true, allowing any unauthenticated visitor to access it. When the ?page=gravitysmtp-settings query parameter is appended, the plugin's register_connector_data() method populates internal connector data, causing the endpoint to return approximately 365 KB of JSON containing the full System Report. This makes it possible for unauthenticated attackers to retrieve detailed system configuration data including PHP version, loaded extensions, web server version, document root path, database server type and version, WordPress version, all active plugins with versions, active theme, WordPress configuration details, database table names, and any API keys/tokens configured in the plugin. |
| CVE-2026-3881 | Medium | 2026-03-31 | Not applicable | Proof of concept only | The Performance Monitor WordPress plugin through 1.0.6 does not validate a parameter before making a request to it, which could allow unauthenticated users to perform SSRF attacks |
| CVE-2026-34887 | Medium | 2026-03-31 | Not applicable | No public exploit | Improper Neutralization of Input During Web Page Generation ('Cross-site Scripting') vulnerability in Extend Themes Kubio AI Page Builder allows Stored XSS.This issue affects Kubio AI Page Builder: from n/a through 2.7.0. |
| CVE-2026-34881 | High | 2026-03-31 | Not applicable | Proof of concept only | OpenStack Glance before 29.1.1, 30.x before 30.1.1, and 31.0.0 is affected by Server-Side Request Forgery (SSRF). By use of HTTP redirects, an authenticated user can bypass URL validation checks and redirect to internal services. Only glance image import functionality is affected. In particular, the web-download and glance-download import methods are subject to this vulnerability, as is the optional (not enabled by default) ovf_process image import plugin. |
| CVE-2026-34784 | High | 2026-03-31 | Not applicable | No public exploit | Parse Server is an open source backend that can be deployed to any infrastructure that can run Node.js. Prior to versions 8.6.71 and 9.7.1-alpha.1, file downloads via HTTP Range requests bypass the afterFind(Parse.File) trigger and its validators on storage adapters that support streaming (e.g. the default GridFS adapter). This allows access to files that should be protected by afterFind trigger authorization logic or built-in validators such as requireUser. This issue has been patched in versions 8.6.71 and 9.7.1-alpha.1. |
| CVE-2026-34740 | Medium | 2026-03-31 | Not applicable | Proof of concept only | WWBN AVideo is an open source video platform. In versions 26.0 and prior, the EPG (Electronic Program Guide) link feature in AVideo allows authenticated users with upload permissions to store arbitrary URLs that the server fetches on every EPG page visit. The URL is validated only with PHP's FILTER_VALIDATE_URL, which accepts internal network addresses. Although AVideo has a dedicated isSSRFSafeURL() function for preventing SSRF, it is not called in this code path. This results in a stored server-side request forgery vulnerability that can be used to scan internal networks, access cloud metadata services, and interact with internal services. At time of publication, there are no publicly available patches. |
| CVE-2026-34739 | Medium | 2026-03-31 | Not applicable | Proof of concept only | WWBN AVideo is an open source video platform. In versions 26.0 and prior, the User_Location plugin's testIP.php page reflects the ip request parameter directly into an HTML input element without applying htmlspecialchars() or any other output encoding. This allows an attacker to inject arbitrary HTML and JavaScript via a crafted URL. Although the page is restricted to admin users, AVideo's SameSite=None cookie configuration allows cross-origin exploitation, meaning an attacker can lure an admin to a malicious link that executes JavaScript in their authenticated session. At time of publication, there are no publicly available patches. |
| CVE-2026-34738 | Medium | 2026-03-31 | Not applicable | Proof of concept only | WWBN AVideo is an open source video platform. In versions 26.0 and prior, AVideo's video processing pipeline accepts an overrideStatus request parameter that allows any uploader to set a video's status to any valid state, including "active" (a). This bypasses the admin-controlled moderation and draft workflows. The setStatus() method validates the status code against a list of known values but does not verify that the caller has permission to set that particular status. As a result, any user with upload permissions can publish videos directly, circumventing content review processes. At time of publication, there are no publicly available patches. |
| CVE-2026-34737 | Medium | 2026-03-31 | Not applicable | Proof of concept only | WWBN AVideo is an open source video platform. In versions 26.0 and prior, the StripeYPT plugin includes a test.php debug endpoint that is accessible to any logged-in user, not just administrators. This endpoint processes Stripe webhook-style payloads and triggers subscription operations, including cancellation. Due to a bug in the retrieveSubscriptions() method that cancels subscriptions instead of merely retrieving them, any authenticated user can cancel arbitrary Stripe subscriptions by providing a subscription ID. At time of publication, there are no publicly available patches. |
| CVE-2026-34733 | High | 2026-03-31 | Not applicable | Proof of concept only | WWBN AVideo is an open source video platform. In versions 26.0 and prior, the AVideo installation script install/deleteSystemdPrivate.php contains a PHP operator precedence bug in its CLI-only access guard. The script is intended to run exclusively from the command line, but the guard condition !php_sapi_name() === 'cli' never evaluates to true due to how PHP resolves operator precedence. The ! (logical NOT) operator binds more tightly than === (strict comparison), causing the expression to always evaluate to false, which means the die() statement never executes. As a result, the script is accessible via HTTP without authentication and will delete files from the server's temp directory while also disclosing the temp directory contents in its response. At time of publication, there are no publicly available patches. |
| CVE-2026-34732 | High | 2026-03-31 | Not applicable | Proof of concept only | WWBN AVideo is an open source video platform. In versions 26.0 and prior, the AVideo CreatePlugin template for list.json.php does not include any authentication or authorization check. While the companion templates add.json.php and delete.json.php both require admin privileges, the list.json.php template was shipped without this guard. Every plugin that uses the CreatePlugin code generator inherits this omission, resulting in 21 unauthenticated data listing endpoints across the platform. These endpoints expose sensitive data including user PII, payment transaction logs, IP addresses, user agents, and internal system records. At time of publication, there are no publicly available patches. |
| CVE-2026-34731 | High | 2026-03-31 | Not applicable | Proof of concept only | WWBN AVideo is an open source video platform. In versions 26.0 and prior, the AVideo on_publish_done.php endpoint in the Live plugin allows unauthenticated users to terminate any active live stream. The endpoint processes RTMP callback events to mark streams as finished in the database, but performs no authentication or authorization checks before doing so. An attacker can enumerate active stream keys from the unauthenticated stats.json.php endpoint, then send crafted POST requests to on_publish_done.php to terminate any live broadcast. This enables denial-of-service against all live streaming functionality on the platform. At time of publication, there are no publicly available patches. |
| CVE-2026-34716 | Medium | 2026-03-31 | Not applicable | Proof of concept only | WWBN AVideo is an open source video platform. In versions 26.0 and prior, the AVideo YPTSocket plugin's caller feature renders incoming call notifications using the jQuery Toast Plugin, passing the caller's display name directly as the heading parameter. The toast plugin constructs the heading as raw HTML ('<h2>' + heading + '</h2>') and inserts it into the DOM via jQuery's .html() method, which parses and executes any embedded HTML or script content. An attacker can set their display name to an XSS payload and trigger code execution on any online user's browser simply by initiating a call - no victim interaction is required beyond being connected to the WebSocket. At time of publication, there are no publicly available patches. |
| CVE-2026-3470 | Low | 2026-03-31 | Outside Waratek scope | No public exploit | A vulnerability exists in the SonicWall Email Security appliance due to improper input sanitization that may lead to data corruption, allowing a remote authenticated attacker as admin user could exploit this issue by providing crafted input that corrupts application database. |
| CVE-2026-3469 | Low | 2026-03-31 | Outside Waratek scope | No public exploit | A denial-of-service (DoS) vulnerability exists due to improper input validation in the SonicWall Email Security appliance, allowing a remote authenticated attacker as admin user to cause the application to become unresponsive. |
| CVE-2026-3468 | Medium | 2026-03-31 | Outside Waratek scope | No public exploit | A stored Cross-Site Scripting (XSS) vulnerability has been identified in the SonicWall Email Security appliance due to improper neutralization of user-supplied input during web page generation, allowing a remote authenticated attacker as admin user to potentially execute arbitrary JavaScript code. |
| CVE-2026-34613 | Medium | 2026-03-31 | Not applicable | Proof of concept only | WWBN AVideo is an open source video platform. In versions 26.0 and prior, the AVideo endpoint objects/pluginSwitch.json.php allows administrators to enable or disable any installed plugin. The endpoint checks for an active admin session but does not validate a CSRF token. Additionally, the plugins database table is explicitly listed in ignoreTableSecurityCheck(), which means the ORM-level Referer/Origin domain validation in ObjectYPT::save() is also bypassed. Combined with SameSite=None on session cookies, an attacker can disable critical security plugins (such as LoginControl for 2FA, subscription enforcement, or access control plugins) by luring an admin to a malicious page. At time of publication, there are no publicly available patches. |
| CVE-2026-34611 | Medium | 2026-03-31 | Not applicable | Proof of concept only | WWBN AVideo is an open source video platform. In versions 26.0 and prior, the AVideo endpoint objects/emailAllUsers.json.php allows administrators to send HTML emails to every registered user on the platform. While the endpoint verifies admin session status, it does not validate a CSRF token. Because AVideo sets SameSite=None on session cookies, a cross-origin POST request from an attacker-controlled page will include the admin's session cookie automatically. An attacker who lures an admin to a malicious page can send an arbitrary HTML email to every user on the platform, appearing to originate from the instance's legitimate SMTP address. At time of publication, there are no publicly available patches. |
| CVE-2026-34605 | High | 2026-03-31 | Not applicable | Working exploit published | SiYuan is a personal knowledge management system. From version 3.6.0 to before version 3.6.2, the SanitizeSVG function introduced in version 3.6.0 to fix XSS in the unauthenticated /api/icon/getDynamicIcon endpoint can be bypassed by using namespace-prefixed element names such as <x:script xmlns:x="http://www.w3.org/2000/svg">. The Go HTML5 parser records the element's tag as "x:script" rather than "script", so the tag check passes it through. The SVG is served with Content-Type: image/svg+xml and no Content Security Policy; when a browser opens the response directly, its XML parser resolves the prefix to the SVG namespace and executes the embedded script. This issue has been patched in version 3.6.2. |
| CVE-2026-34595 | Medium | 2026-03-31 | Not applicable | No public exploit | Parse Server is an open source backend that can be deployed to any infrastructure that can run Node.js. Prior to versions 8.6.70 and 9.7.0-alpha.18, an authenticated user with find class-level permission can bypass the protectedFields class-level permission setting on LiveQuery subscriptions. By sending a subscription with a $or, $and, or $nor operator value as a plain object with numeric keys and a length property (an "array-like" object) instead of an array, the protected-field guard is bypassed. The subscription event firing acts as a binary oracle, allowing the attacker to infer whether a protected field matches a given test value. This issue has been patched in versions 8.6.70 and 9.7.0-alpha.18. |
| CVE-2026-34586 | Medium | 2026-03-31 | Not applicable | Proof of concept only | PdfDing is a selfhosted PDF manager, viewer and editor offering a seamless user experience on multiple devices. Prior to version 1.7.1, check_shared_access_allowed() validates only session existence — it does not check SharedPdf.inactive (expiration / max views) or SharedPdf.deleted. The Serve and Download endpoints rely solely on this function, allowing previously-authorized users to access shared PDF content after expiration, view limit, or soft-deletion. This issue has been patched in version 1.7.1. |
| CVE-2026-34585 | High | 2026-03-31 | Not applicable | Proof of concept only | SiYuan is a personal knowledge management system. Prior to version 3.6.2, a vulnerability allows crafted block attribute values to bypass server-side attribute escaping when an HTML entity is mixed with raw special characters. An attacker can embed a malicious IAL value inside a .sy document, package it as a .sy.zip, and have the victim import it through the normal Import -> SiYuan .sy.zip workflow. Once the note is opened, the malicious attribute breaks out of its original HTML context and injects an event handler, resulting in stored XSS. In the Electron desktop client, this XSS reaches remote code execution because injected JavaScript runs with access to Node/Electron APIs. This issue has been patched in version 3.6.2. |
| CVE-2026-34574 | Medium | 2026-03-31 | Not applicable | No public exploit | Parse Server is an open source backend that can be deployed to any infrastructure that can run Node.js. Prior to versions 8.6.69 and 9.7.0-alpha.14, an authenticated user can bypass the immutability guard on session fields (expiresAt, createdWith) by sending a null value in a PUT request to the session update endpoint. This allows nullifying the session expiry, making the session valid indefinitely and bypassing configured session length policies. This issue has been patched in versions 8.6.69 and 9.7.0-alpha.14. |
| CVE-2026-34573 | High | 2026-03-31 | Not applicable | No public exploit | Parse Server is an open source backend that can be deployed to any infrastructure that can run Node.js. Prior to versions 8.6.68 and 9.7.0-alpha.12, the GraphQL query complexity validator can be exploited to cause a denial-of-service by sending a crafted query with binary fan-out fragment spreads. A single unauthenticated request can block the Node.js event loop for seconds, denying service to all concurrent users. This only affects deployments that have enabled the requestComplexity.graphQLDepth or requestComplexity.graphQLFields configuration options. This issue has been patched in versions 8.6.68 and 9.7.0-alpha.12. |
| CVE-2026-34556 | Medium | 2026-03-31 | Not applicable | Proof of concept only | iccDEV provides a set of libraries and tools for working with ICC color management profiles. Prior to version 2.3.1.6, there is a heap-buffer-overflow (HBO) in icAnsiToUtf8() in the XML conversion path. The issue is triggered by a crafted ICC profile which causes icAnsiToUtf8(std::string&, char const*) to treat an input buffer as a C-string and call operations that rely on strlen()/null-termination. AddressSanitizer reports an out-of-bounds READ of size 115 past a 114-byte heap allocation, with the failure observed while running the iccToXml tool. This issue has been patched in version 2.3.1.6. |
| CVE-2026-34555 | Medium | 2026-03-31 | Not applicable | Proof of concept only | iccDEV provides a set of libraries and tools for working with ICC color management profiles. Prior to version 2.3.1.6, there is a stack-buffer-overflow (SBO) in CIccTagFixedNum<>::GetValues() and a related bug chain. The primary crash is an AddressSanitizer-reported WRITE of size 4 that overflows a 4-byte stack variable (rv) via the call chain CIccTagFixedNum::GetValues() -> CIccTagStruct::GetElemNumberValue(). This issue has been patched in version 2.3.1.6. |
| CVE-2026-34554 | Medium | 2026-03-31 | Not applicable | No public exploit | iccDEV provides a set of libraries and tools for working with ICC color management profiles. Prior to version 2.3.1.6, a heap-buffer-overflow (HBO) in CIccApplyCmmSearch::costFunc() can be triggered via malformed JSON configuration input to the iccApplySearch tool. AddressSanitizer reports an out-of-bounds READ of size 8 originating from CIccApplyCmmSearch::costFunc(CIccSearchVec&) at IccProfLib/IccCmmSearch.cpp:112:5. This issue has been patched in version 2.3.1.6. |
| CVE-2026-34553 | Medium | 2026-03-31 | Not applicable | No public exploit | iccDEV provides a set of libraries and tools for working with ICC color management profiles. Prior to version 2.3.1.6, there is a defect in LUT dump/iteration logic affecting CIccCLUT::Iterate() and output produced by CIccMBB::Describe() (via CLUT dumping). This issue has been patched in version 2.3.1.6. |
| CVE-2026-34552 | Medium | 2026-03-31 | Not applicable | Proof of concept only | iccDEV provides a set of libraries and tools for working with ICC color management profiles. Prior to version 2.3.1.6, there is an Undefined Behavior (UB) issue in IccTagLut.cpp where the code performs member access through a null pointer of type CIccApplyCLUT. This issue has been patched in version 2.3.1.6. |
| CVE-2026-34551 | Medium | 2026-03-31 | Not applicable | Proof of concept only | iccDEV provides a set of libraries and tools for working with ICC color management profiles. Prior to version 2.3.1.6, a null-pointer dereference (NPD) in CIccTagLut16::Write() can be triggered when processing a crafted ICC profile (embedded in a TIFF and extracted during iccTiffDump). This issue has been patched in version 2.3.1.6. |
| CVE-2026-34550 | Medium | 2026-03-31 | Not applicable | No public exploit | iccDEV provides a set of libraries and tools for working with ICC color management profiles. Prior to version 2.3.1.6, there is an Undefined Behavior (UB) condition in IccProfLib/IccIO.cpp caused by an implicit conversion from a negative signed integer to size_t (unsigned), which changes the value. This issue has been patched in version 2.3.1.6. |
| CVE-2026-34549 | Medium | 2026-03-31 | Not applicable | Proof of concept only | iccDEV provides a set of libraries and tools for working with ICC color management profiles. Prior to version 2.3.1.6, there is an Undefined Behavior (UB) condition in IccUtil.cpp triggered by a crafted input profile. Under UndefinedBehaviorSanitizer, the issue is reported as invalid left shift operations on icUInt32Number (unsigned 32-bit) where the shifted value “cannot be represented” in that type. This issue has been patched in version 2.3.1.6. |
| CVE-2026-34548 | Medium | 2026-03-31 | Not applicable | No public exploit | iccDEV provides a set of libraries and tools for working with ICC color management profiles. Prior to version 2.3.1.6, there is an Undefined Behavior (UB) condition in the XML conversion tooling path (iccToXml) caused by an implicit conversion from a negative signed integer to icUInt32Number (unsigned 32-bit), which changes the value. This issue has been patched in version 2.3.1.6. |
| CVE-2026-34547 | Medium | 2026-03-31 | Not applicable | Proof of concept only | iccDEV provides a set of libraries and tools for working with ICC color management profiles. Prior to version 2.3.1.6, an Undefined Behavior (UB) condition in IccUtil.cpp can be triggered by a crafted ICC profile when running iccDumpProfile. This issue has been patched in version 2.3.1.6. |
| CVE-2026-34546 | Medium | 2026-03-31 | Not applicable | No public exploit | iccDEV provides a set of libraries and tools for working with ICC color management profiles. Prior to version 2.3.1.6, a crafted TIFF input can trigger Undefined Behavior (UB) due to division by zero in the TIFF handling code paths used by iccTiffDump. This issue has been patched in version 2.3.1.6. |
| CVE-2026-34542 | Medium | 2026-03-31 | Not applicable | Proof of concept only | iccDEV provides a set of libraries and tools for working with ICC color management profiles. Prior to version 2.3.1.6, a crafted ICC profile can trigger a stack-buffer-overflow (SBO) in CIccCalculatorFunc::Apply() when processed via iccApplyNamedCmm. Under AddressSanitizer, the failure is reported as a 4-byte write stack-buffer-overflow in IccProfLib/IccMpeCalc.cpp:3873, reachable through the MPE calculator / curve set initialization path. This issue has been patched in version 2.3.1.6. |
| CVE-2026-34541 | Medium | 2026-03-31 | Not applicable | Proof of concept only | iccDEV provides a set of libraries and tools for working with ICC color management profiles. Prior to version 2.3.1.6, a crafted ICC profile can trigger Undefined Behavior (UB) via a null-pointer member call in CIccCombinedConnectionConditions::CIccCombinedConnectionConditions() (reported by UBSan as “member call on null pointer of type CIccTagSpectralViewingConditions”). The issue is reachable when running iccApplyNamedCmm with -PCC using a malformed .icc profile. This issue has been patched in version 2.3.1.6. |
| CVE-2026-34540 | Medium | 2026-03-31 | Not applicable | Proof of concept only | iccDEV provides a set of libraries and tools for working with ICC color management profiles. Prior to version 2.3.1.6, a crafted ICC profile can trigger a heap-buffer-overflow (HBO) in icMemDump() when iccDumpProfile attempts to dump/describe malformed tag contents. The issue is observable under AddressSanitizer as an out-of-bounds heap read in icMemDump(...) at IccProfLib/IccUtil.cpp:1002, reachable via CIccTagUnknown::Describe(). This issue has been patched in version 2.3.1.6. |
| CVE-2026-34539 | Medium | 2026-03-31 | Not applicable | No public exploit | iccDEV provides a set of libraries and tools for working with ICC color management profiles. Prior to version 2.3.1.6, a crafted ICC profile and TIFF input can trigger a heap-buffer-overflow (HBO) in CTiffImg::WriteLine(). The issue is observable under AddressSanitizer as an out-of-bounds heap read when running iccSpecSepToTiff on a malicious .icc + .tif pair, leading to a crash during TIFF strip writing. This issue has been patched in version 2.3.1.6. |
| CVE-2026-34537 | Medium | 2026-03-31 | Not applicable | Proof of concept only | iccDEV provides a set of libraries and tools for working with ICC color management profiles. Prior to version 2.3.1.6, a crafted ICC profile can trigger Undefined Behavior (UB) in CIccOpDefEnvVar::Exec() due to invalid enum values being loaded for icSigCmmEnvVar. The issue is observable under UBSan as a “load of value … not a valid value for type icSigCmmEnvVar”, indicating an invalid enum/type value being consumed during ICC profile processing. This issue has been patched in version 2.3.1.6. |
| CVE-2026-34536 | Medium | 2026-03-31 | Not applicable | No public exploit | iccDEV provides a set of libraries and tools for working with ICC color management profiles. Prior to version 2.3.1.6, a crafted ICC profile can trigger a stack overflow (SO) in SIccCalcOp::ArgsUsed(). The issue is observable under AddressSanitizer as a stack-overflow when iccApplyProfiles processes a malicious profile, with the crash occurring while computing argument usage during calculator underflow/overflow checks. This issue has been patched in version 2.3.1.6. |
| CVE-2026-34535 | Medium | 2026-03-31 | Not applicable | Proof of concept only | iccDEV provides a set of libraries and tools for working with ICC color management profiles. Prior to version 2.3.1.6, a crafted ICC profile can trigger a segmentation fault (SEGV) in CIccTagArray::Cleanup(). The issue is observable under UBSan/ASan as misaligned member access / misaligned pointer loads followed by an invalid read leading to process crash when running iccRoundTrip on a malicious profile. This issue has been patched in version 2.3.1.6. |
| CVE-2026-34534 | Medium | 2026-03-31 | Not applicable | No public exploit | iccDEV provides a set of libraries and tools for working with ICC color management profiles. Prior to version 2.3.1.6, a crafted ICC profile can trigger a heap-buffer-overflow (HBO) in CIccMpeSpectralMatrix::Describe(). The issue is observable under AddressSanitizer as an out-of-bounds heap read when running iccDumpProfile on a malicious profile. This issue has been patched in version 2.3.1.6. |
| CVE-2026-34533 | Medium | 2026-03-31 | Not applicable | Proof of concept only | iccDEV provides a set of libraries and tools for working with ICC color management profiles. Prior to version 2.3.1.6, a crafted ICC profile can trigger Undefined Behavior (UB) in CIccCalculatorFunc::ApplySequence() due to invalid enum values being loaded for icChannelFuncSignature. The issue is observable under UBSan as a “load of value … not a valid value for type icChannelFuncSignature”, indicating a type/enum value confusion scenario during ICC profile processing. This issue has been patched in version 2.3.1.6. |
| CVE-2026-34532 | Critical | 2026-03-31 | Not applicable | No public exploit | Parse Server is an open source backend that can be deployed to any infrastructure that can run Node.js. Prior to versions 8.6.67 and 9.7.0-alpha.11, an attacker can bypass Cloud Function validator access controls by appending "prototype.constructor" to the function name in the URL. When a Cloud Function handler is declared using the function keyword and its validator is a plain object or arrow function, the trigger store traversal resolves the handler through its own prototype chain while the validator store fails to mirror this traversal, causing all access control enforcement to be skipped. This allows unauthenticated callers to invoke Cloud Functions that are meant to be protected by validators such as requireUser, requireMaster, or custom validation logic. This issue has been patched in versions 8.6.67 and 9.7.0-alpha.11. |
| CVE-2026-34509 | Unknown | 2026-03-31 | Withdrawn CVE | No public exploit | Rejected reason: This CVE ID has been rejected or withdrawn by its CVE Numbering Authority. |
| CVE-2026-34508 | Unknown | 2026-03-31 | Withdrawn CVE | No public exploit | Rejected reason: This CVE ID has been rejected or withdrawn by its CVE Numbering Authority. |
| CVE-2026-34506 | Low | 2026-03-31 | Not applicable | No public exploit | OpenClaw before 2026.3.8 contains a sender allowlist bypass vulnerability in its Microsoft Teams plugin that allows unauthorized senders to bypass intended authorization checks. When a team/channel route allowlist is configured with an empty groupAllowFrom parameter, the message handler synthesizes wildcard sender authorization, permitting any sender in the matched team/channel to trigger replies in allowlisted Teams routes. |
| CVE-2026-34505 | Medium | 2026-03-31 | Not applicable | No public exploit | OpenClaw before 2026.3.12 applies rate limiting only after successful webhook authentication, allowing attackers to bypass rate limits and brute-force webhook secrets. Attackers can submit repeated authentication requests with invalid secrets without triggering rate limit responses, enabling systematic secret guessing and subsequent forged webhook submission. |
| CVE-2026-34504 | Medium | 2026-03-31 | Not applicable | No public exploit | OpenClaw before 2026.3.28 contains a server-side request forgery vulnerability in the fal provider image-generation-provider.ts component that allows attackers to fetch internal URLs. A malicious or compromised fal relay can exploit unguarded image download fetches to expose internal service metadata and responses through the image pipeline. |
| CVE-2026-34503 | High | 2026-03-31 | Not applicable | No public exploit | OpenClaw before 2026.3.28 fails to disconnect active WebSocket sessions when devices are removed or tokens are revoked. Attackers with revoked credentials can maintain unauthorized access through existing live sessions until forced reconnection. |
| CVE-2026-34453 | High | 2026-03-31 | Not applicable | Working exploit published | SiYuan is a personal knowledge management system. Prior to version 3.6.2, the publish service exposes bookmarked blocks from password-protected documents to unauthenticated visitors. In publish/read-only mode, /api/bookmark/getBookmark filters bookmark results by calling FilterBlocksByPublishAccess(nil, ...). Because the filter treats a nil context as authorized, it skips the publish password check and returns bookmarked blocks from documents configured as Protected. As a result, anyone who can access the publish service can retrieve content from protected documents without providing the required password, as long as at least one block in the document is bookmarked. This issue has been patched in version 3.6.2. |
| CVE-2026-34452 | Medium | 2026-03-31 | Not applicable | No public exploit | The Claude SDK for Python provides access to the Claude API from Python applications. From version 0.86.0 to before version 0.87.0, the async local filesystem memory tool in the Anthropic Python SDK validated that model-supplied paths resolved inside the sandboxed memory directory, but then returned the unresolved path for subsequent file operations. A local attacker able to write to the memory directory could retarget a symlink between validation and use, causing reads or writes to escape the sandbox. The synchronous memory tool implementation was not affected. This issue has been patched in version 0.87.0. |
| CVE-2026-34451 | Medium | 2026-03-31 | Not applicable | No public exploit | Claude SDK for TypeScript provides access to the Claude API from server-side TypeScript or JavaScript applications. From version 0.79.0 to before version 0.81.0, the local filesystem memory tool in the Anthropic TypeScript SDK validated model-supplied paths using a string prefix check that did not append a trailing path separator. A model steered by prompt injection could supply a crafted path that resolved to a sibling directory sharing the memory root's name as a prefix, allowing reads and writes outside the sandboxed memory directory. This issue has been patched in version 0.81.0. |
| CVE-2026-34450 | Medium | 2026-03-31 | Not applicable | No public exploit | The Claude SDK for Python provides access to the Claude API from Python applications. From version 0.86.0 to before version 0.87.0, the local filesystem memory tool in the Anthropic Python SDK created memory files with mode 0o666, leaving them world-readable on systems with a standard umask and world-writable in environments with a permissive umask such as many Docker base images. A local attacker on a shared host could read persisted agent state, and in containerized deployments could modify memory files to influence subsequent model behavior. Both the synchronous and asynchronous memory tool implementations were affected. This issue has been patched in version 0.87.0. |