CWE-349

Next.js is a React framework for building full-stack web applications. By sending a crafted HTTP request, it is possible to poison the cache of a non-dynamic server-side rendered route in the pages router (this does not affect the app router). When this crafted request is sent it could coerce Next.js to cache a route that is meant to not be cached and send a `Cache-Control: s-maxage=1, stale-while-revalidate` header which some upstream CDNs may cache as well. To be potentially affected all of the following must apply: 1. Next.js between 13.5.1 and 14.2.9, 2. Using pages router, & 3. Using non-dynamic server-side rendered routes e.g. `pages/dashboard.tsx` not `pages/blog/[slug].tsx`. This vulnerability was resolved in Next.js v13.5.7, v14.2.10, and later. We recommend upgrading regardless of whether you can reproduce the issue or not. There are no official or recommended workarounds for this issue, we recommend that users patch to a safe version.

An issue was discovered in Http Foundation in Symfony 2.7.0 through 2.7.48, 2.8.0 through 2.8.43, 3.3.0 through 3.3.17, 3.4.0 through 3.4.13, 4.0.0 through 4.0.13, and 4.1.0 through 4.1.2. It arises from support for a (legacy) IIS header that lets users override the path in the request URL via the X-Original-URL or X-Rewrite-URL HTTP request header. These headers are designed for IIS support, but it's not verified that the server is in fact running IIS, which means anybody who can send these requests to an application can trigger this. This affects \Symfony\Component\HttpFoundation\Request::prepareRequestUri() where X-Original-URL and X_REWRITE_URL are both used. The fix drops support for these methods so that they cannot be used as attack vectors such as web cache poisoning.

Django 1.4 before 1.4.13, 1.5 before 1.5.8, 1.6 before 1.6.5, and 1.7 before 1.7b4 does not properly include the (1) Vary: Cookie or (2) Cache-Control header in responses, which allows remote attackers to obtain sensitive information or poison the cache via a request from certain browsers.

A vulnerability exists in the way that iTerm2 integrates with tmux's control mode, which may allow an attacker to execute arbitrary commands by providing malicious output to the terminal. This affects versions of iTerm2 up to and including 3.3.5. This vulnerability may allow an attacker to execute arbitrary commands on their victim's computer by providing malicious output to the terminal. It could be exploited using command-line utilities that print attacker-controlled content.

Django before 1.2.7 and 1.3.x before 1.3.1 uses a request's HTTP Host header to construct a full URL in certain circumstances, which allows remote attackers to conduct cache poisoning attacks via a crafted request.

Acceptance of extraneous untrusted data with trusted data in Windows COM allows an unauthorized attacker to elevate privileges locally.

Infinispan permits improper deserialization of trusted data via XML and JSON transcoders under certain server configurations. A user with authenticated access to the server could send a malicious object to a cache configured to accept certain types of objects, achieving code execution and possible further attacks. Versions 9.0.3.Final, 9.1.7.Final, 8.2.10.Final, 9.2.2.Final, 9.3.0.Alpha1 are believed to be affected.

Nimble is a package manager for the Nim programming language. In Nim release versions before versions 1.2.10 and 1.4.4, "nimble refresh" fetches a list of Nimble packages over HTTPS without full verification of the SSL/TLS certificate due to the default setting of httpClient. An attacker able to perform MitM can deliver a modified package list containing malicious software packages. If the packages are installed and used the attack escalates to untrusted code execution.

An issue in AsyncSSH before 2.14.1 allows attackers to control the remote end of an SSH client session via packet injection/removal and shell emulation, aka a "Rogue Session Attack."

Vulnerability in the Oracle Java SE, Oracle GraalVM for JDK, Oracle GraalVM Enterprise Edition product of Oracle Java SE (component: Hotspot). Supported versions that are affected are Oracle Java SE: 8u401, 8u401-perf, 11.0.22, 17.0.10, 21.0.2, 22; Oracle GraalVM for JDK: 17.0.10, 21.0.2, 22; Oracle GraalVM Enterprise Edition: 20.3.13 and 21.3.9. Difficult to exploit vulnerability allows unauthenticated attacker with network access via multiple protocols to compromise Oracle Java SE, Oracle GraalVM for JDK, Oracle GraalVM Enterprise Edition. Successful attacks of this vulnerability can result in unauthorized update, insert or delete access to some of Oracle Java SE, Oracle GraalVM for JDK, Oracle GraalVM Enterprise Edition accessible data. Note: This vulnerability can be exploited by using APIs in the specified Component, e.g., through a web service which supplies data to the APIs. This vulnerability also applies to Java deployments, typically in clients running sandboxed Java Web Start applications or sandboxed Java applets, that load and run untrusted code (e.g., code that comes from the internet) and rely on the Java sandbox for security. CVSS 3.1 Base Score 3.7 (Integrity impacts). CVSS Vector: (CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N).

An issue in AsyncSSH before 2.14.1 allows attackers to control the extension info message (RFC 8308) via a man-in-the-middle attack, aka a "Rogue Extension Negotiation."

Under certain circumstances, BIND is too lenient when accepting records from answers, allowing an attacker to inject forged data into the cache. This issue affects BIND 9 versions 9.11.0 through 9.16.50, 9.18.0 through 9.18.39, 9.20.0 through 9.20.13, 9.21.0 through 9.21.12, 9.11.3-S1 through 9.16.50-S1, 9.18.11-S1 through 9.18.39-S1, and 9.20.9-S1 through 9.20.13-S1.

Acceptance of extraneous untrusted data with trusted data in Windows BitLocker allows an unauthorized attacker to bypass a security feature with a physical attack.

A acceptance of Extraneous Untrusted Data With Trusted Data vulnerability in the start script of openldap2 of SUSE Enterprise Storage 5, SUSE Linux Enterprise Debuginfo 11-SP3, SUSE Linux Enterprise Debuginfo 11-SP4, SUSE Linux Enterprise Point of Sale 11-SP3, SUSE Linux Enterprise Server 11-SECURITY, SUSE Linux Enterprise Server 11-SP4-LTSS, SUSE Linux Enterprise Server 12-SP2-BCL, SUSE Linux Enterprise Server 12-SP2-LTSS, SUSE Linux Enterprise Server 12-SP3-BCL, SUSE Linux Enterprise Server 12-SP3-LTSS, SUSE Linux Enterprise Server 12-SP4, SUSE Linux Enterprise Server 12-SP5, SUSE Linux Enterprise Server 15-LTSS, SUSE Linux Enterprise Server for SAP 12-SP2, SUSE Linux Enterprise Server for SAP 12-SP3, SUSE Linux Enterprise Server for SAP 15, SUSE OpenStack Cloud 7, SUSE OpenStack Cloud 8, SUSE OpenStack Cloud Crowbar 8; openSUSE Leap 15.1, openSUSE Leap 15.2 allows local attackers to escalate privileges from user ldap to root. This issue affects: SUSE Enterprise Storage 5 openldap2 versions prior to 2.4.41-18.71.2. SUSE Linux Enterprise Debuginfo 11-SP3 openldap2 versions prior to 2.4.26-0.74.13.1,. SUSE Linux Enterprise Debuginfo 11-SP4 openldap2 versions prior to 2.4.26-0.74.13.1,. SUSE Linux Enterprise Point of Sale 11-SP3 openldap2 versions prior to 2.4.26-0.74.13.1,. SUSE Linux Enterprise Server 11-SECURITY openldap2-client-openssl1 versions prior to 2.4.26-0.74.13.1. SUSE Linux Enterprise Server 11-SP4-LTSS openldap2 versions prior to 2.4.26-0.74.13.1,. SUSE Linux Enterprise Server 12-SP2-BCL openldap2 versions prior to 2.4.41-18.71.2. SUSE Linux Enterprise Server 12-SP2-LTSS openldap2 versions prior to 2.4.41-18.71.2. SUSE Linux Enterprise Server 12-SP3-BCL openldap2 versions prior to 2.4.41-18.71.2. SUSE Linux Enterprise Server 12-SP3-LTSS openldap2 versions prior to 2.4.41-18.71.2. SUSE Linux Enterprise Server 12-SP4 openldap2 versions prior to 2.4.41-18.71.2. SUSE Linux Enterprise Server 12-SP5 openldap2 versions prior to 2.4.41-18.71.2. SUSE Linux Enterprise Server 15-LTSS openldap2 versions prior to 2.4.46-9.31.1. SUSE Linux Enterprise Server for SAP 12-SP2 openldap2 versions prior to 2.4.41-18.71.2. SUSE Linux Enterprise Server for SAP 12-SP3 openldap2 versions prior to 2.4.41-18.71.2. SUSE Linux Enterprise Server for SAP 15 openldap2 versions prior to 2.4.46-9.31.1. SUSE OpenStack Cloud 7 openldap2 versions prior to 2.4.41-18.71.2. SUSE OpenStack Cloud 8 openldap2 versions prior to 2.4.41-18.71.2. SUSE OpenStack Cloud Crowbar 8 openldap2 versions prior to 2.4.41-18.71.2. openSUSE Leap 15.1 openldap2 versions prior to 2.4.46-lp151.10.12.1. openSUSE Leap 15.2 openldap2 versions prior to 2.4.46-lp152.14.3.1.

Improper input validation in Microsoft Office Word allows an unauthorized attacker to bypass a security feature over a network.

dnsjava is an implementation of DNS in Java. Records in DNS replies are not checked for their relevance to the query, allowing an attacker to respond with RRs from different zones. This vulnerability is fixed in 3.6.0.

A vulnerability has been identified in RUGGEDCOM RM1224 LTE(4G) EU (6GK6108-4AM00-2BA2) (All versions < V7.2.2), RUGGEDCOM RM1224 LTE(4G) NAM (6GK6108-4AM00-2DA2) (All versions < V7.2.2), SCALANCE M804PB (6GK5804-0AP00-2AA2) (All versions < V7.2.2), SCALANCE M812-1 ADSL-Router (6GK5812-1AA00-2AA2) (All versions < V7.2.2), SCALANCE M812-1 ADSL-Router (6GK5812-1BA00-2AA2) (All versions < V7.2.2), SCALANCE M816-1 ADSL-Router (6GK5816-1AA00-2AA2) (All versions < V7.2.2), SCALANCE M816-1 ADSL-Router (6GK5816-1BA00-2AA2) (All versions < V7.2.2), SCALANCE M826-2 SHDSL-Router (6GK5826-2AB00-2AB2) (All versions < V7.2.2), SCALANCE M874-2 (6GK5874-2AA00-2AA2) (All versions < V7.2.2), SCALANCE M874-3 (6GK5874-3AA00-2AA2) (All versions < V7.2.2), SCALANCE M876-3 (6GK5876-3AA02-2BA2) (All versions < V7.2.2), SCALANCE M876-3 (ROK) (6GK5876-3AA02-2EA2) (All versions < V7.2.2), SCALANCE M876-4 (6GK5876-4AA10-2BA2) (All versions < V7.2.2), SCALANCE M876-4 (EU) (6GK5876-4AA00-2BA2) (All versions < V7.2.2), SCALANCE M876-4 (NAM) (6GK5876-4AA00-2DA2) (All versions < V7.2.2), SCALANCE MUM853-1 (EU) (6GK5853-2EA00-2DA1) (All versions < V7.2.2), SCALANCE MUM856-1 (EU) (6GK5856-2EA00-3DA1) (All versions < V7.2.2), SCALANCE MUM856-1 (RoW) (6GK5856-2EA00-3AA1) (All versions < V7.2.2), SCALANCE S615 EEC LAN-Router (6GK5615-0AA01-2AA2) (All versions < V7.2.2), SCALANCE S615 LAN-Router (6GK5615-0AA00-2AA2) (All versions < V7.2.2), SCALANCE WAB762-1 (6GK5762-1AJ00-6AA0) (All versions < V3.0.0), SCALANCE WAM763-1 (6GK5763-1AL00-7DA0) (All versions < V3.0.0), SCALANCE WAM763-1 (ME) (6GK5763-1AL00-7DC0) (All versions < V3.0.0), SCALANCE WAM763-1 (US) (6GK5763-1AL00-7DB0) (All versions < V3.0.0), SCALANCE WAM766-1 (6GK5766-1GE00-7DA0) (All versions < V3.0.0), SCALANCE WAM766-1 (ME) (6GK5766-1GE00-7DC0) (All versions < V3.0.0), SCALANCE WAM766-1 (US) (6GK5766-1GE00-7DB0) (All versions < V3.0.0), SCALANCE WAM766-1 EEC (6GK5766-1GE00-7TA0) (All versions < V3.0.0), SCALANCE WAM766-1 EEC (ME) (6GK5766-1GE00-7TC0) (All versions < V3.0.0), SCALANCE WAM766-1 EEC (US) (6GK5766-1GE00-7TB0) (All versions < V3.0.0), SCALANCE WUB762-1 (6GK5762-1AJ00-1AA0) (All versions < V3.0.0), SCALANCE WUB762-1 iFeatures (6GK5762-1AJ00-2AA0) (All versions < V3.0.0), SCALANCE WUM763-1 (6GK5763-1AL00-3AA0) (All versions < V3.0.0), SCALANCE WUM763-1 (6GK5763-1AL00-3DA0) (All versions < V3.0.0), SCALANCE WUM763-1 (US) (6GK5763-1AL00-3AB0) (All versions < V3.0.0), SCALANCE WUM763-1 (US) (6GK5763-1AL00-3DB0) (All versions < V3.0.0), SCALANCE WUM766-1 (6GK5766-1GE00-3DA0) (All versions < V3.0.0), SCALANCE WUM766-1 (ME) (6GK5766-1GE00-3DC0) (All versions < V3.0.0), SCALANCE WUM766-1 (USA) (6GK5766-1GE00-3DB0) (All versions < V3.0.0). Affected products do not properly validate the content of uploaded X509 certificates which could allow an attacker with administrative privileges to execute arbitrary code on the device.

Acceptance of extraneous untrusted data with trusted data in UrlMon allows an unauthorized attacker to bypass a security feature over a network.

Nuxt is an open-source web development framework for Vue.js. Prior to 3.16.0, by sending a crafted HTTP request to a server behind an CDN, it is possible in some circumstances to poison the CDN cache and highly impacts the availability of a site. It is possible to craft a request, such as https://mysite.com/?/_payload.json which will be rendered as JSON. If the CDN in front of a Nuxt site ignores the query string when determining whether to cache a route, then this JSON response could be served to future visitors to the site. An attacker can perform this attack to a vulnerable site in order to make a site unavailable indefinitely. It is also possible in the case where the cache will be reset to make a small script to send a request each X seconds (=caching duration) so that the cache is permanently poisoned making the site completely unavailable. This vulnerability is fixed in 3.16.0.

A flaw was found in the Linux kernels SELinux LSM hook implementation before version 5.7, where it incorrectly assumed that an skb would only contain a single netlink message. The hook would incorrectly only validate the first netlink message in the skb and allow or deny the rest of the messages within the skb with the granted permission without further processing.

A vulnerability exists in NGINX OSS and NGINX Plus when configured to proxy to upstream Transport Layer Security (TLS) servers. An attacker with a man-in-the-middle (MITM) position on the upstream server side—along with conditions beyond the attacker's control—may be able to inject plain text data into the response from an upstream proxied server.  Note: Software versions which have reached End of Technical Support (EoTS) are not evaluated.

In JetBrains IntelliJ IDEA before 2023.3.2 code execution was possible in Untrusted Project mode via a malicious plugin repository specified in the project configuration

NLnet Labs Unbound up to and including version 1.24.1 is vulnerable to possible domain hijack attacks. Promiscuous NS RRSets that complement positive DNS replies in the authority section can be used to trick resolvers to update their delegation information for the zone. Usually these RRSets are used to update the resolver's knowledge of the zone's name servers. A malicious actor can exploit the possible poisonous effect by injecting NS RRSets (and possibly their respective address records) in a reply. This could be done for example by trying to spoof a packet or fragmentation attacks. Unbound would then proceed to update the NS RRSet data it already has since the new data has enough trust for it, i.e., in-zone data for the delegation point. Unbound 1.24.1 includes a fix that scrubs unsolicited NS RRSets (and their respective address records) from replies mitigating the possible poison effect. Unbound 1.24.2 includes an additional fix that scrubs unsolicited NS RRSets (and their respective address records) from YXDOMAIN and non-referral nodata replies, further mitigating the possible poison effect.

ESP-NOW Component provides a connectionless Wi-Fi communication protocol. An replay attacks vulnerability was discovered in the implementation of the ESP-NOW because the caches is not differentiated by message types, it is a single, shared resource for all kinds of messages, whether they are broadcast or unicast, and regardless of whether they are ciphertext or plaintext. This can result an attacker to clear the cache of its legitimate entries, there by creating an opportunity to re-inject previously captured packets. This vulnerability is fixed in 2.5.2.

Stronger revision number limitations were required on file serving endpoints to improve cache poisoning protection.

OpenClaw before 2026.4.2 contains an insufficient scope vulnerability in Zalo webhook replay dedupe keys that allows legitimate events from different conversations or senders to collide. Attackers can exploit weak deduplication scoping to cause silent message suppression and disrupt bot workflows across chat sessions.

Acceptance of extraneous untrusted data with trusted data vulnerability exists in EC-CUBE 4 series. If this vulnerability is exploited, an attacker who obtained the administrative privilege may install an arbitrary PHP package. If the obsolete versions of PHP packages are installed, the product may be affected by some known vulnerabilities.

No cwe for this issue in Windows DHCP Server allows an unauthorized attacker to perform tampering over a network.

NLnet Labs Unbound up to and including version 1.25.0 is vulnerable to poisoning via promiscuous records for the authority section. Promiscuous RRSets that complement DNS replies in the authority section can be used to trick Unbound to cache such records. If an adversary is able to attach such records in a reply (i.e., spoofed packet, fragmentation attack) he would be able to poison Unbound's cache. A malicious actor can exploit the possible poisonous effect by injecting RRSets other than NS that are also accompanied by address records in a reply, for example MX. This could be achieved by trying to spoof a reply packet or fragmentation attacks. Unbound would then accept the relative address records in the additional section and cache them if the authority RRSet has enough trust at this point, i.e., in-zone data for the delegation point. Unbound 1.25.1 contains a patch with a fix that disregards address records from the additional section if they are not explicitly relevant only to authority NS records, mitigating the possible poison effect. This is a complement fix to CVE-2025-11411.

A `named` caching resolver that is configured to send ECS (EDNS Client Subnet) options may be vulnerable to a cache-poisoning attack. This issue affects BIND 9 versions 9.11.3-S1 through 9.16.50-S1, 9.18.11-S1 through 9.18.37-S1, and 9.20.9-S1 through 9.20.10-S1.

aiosmptd is a reimplementation of the Python stdlib smtpd.py based on asyncio. Prior to version 1.4.6, servers based on aiosmtpd accept extra unencrypted commands after STARTTLS, treating them as if they came from inside the encrypted connection. This could be exploited by a man-in-the-middle attack. Version 1.4.6 contains a patch for the issue.

The WP Go Maps (formerly WP Google Maps) plugin for WordPress is vulnerable to Cache Poisoning in all versions up to, and including, 9.0.48. This is due to the plugin not serving cached data from server-side responses and instead relying on user-input. This makes it possible for unauthenticated attackers to poison the cache location for location search results.

Next.js is a React framework for building full-stack web applications. From 12.2.0 to before 15.5.16 and 16.2.5, an external client could send a x-nextjs-data header on a normal request to a path handled by middleware that returns a redirect. When that happened, the middleware/proxy could treat the request as a data request and replace the standard Location redirect header with the internal x-nextjs-redirect header. Browsers do not follow x-nextjs-redirect, so the response became an unusable redirect for normal clients. If the application was deployed behind a CDN or reverse proxy that caches 3xx responses without varying on this header, a single attacker request could poison the cached redirect response for the affected path. Subsequent visitors could then receive a cached redirect response without a Location header, causing a denial of service for that redirect path until the cache entry expired or was purged. This vulnerability is fixed in 15.5.16 and 16.2.5.

A multi-vendor cache poisoning vulnerability named 'Rebirthday Attack' has been discovered in caching resolvers that support EDNS Client Subnet (ECS). Unbound is also vulnerable when compiled with ECS support, i.e., '--enable-subnet', AND configured to send ECS information along with queries to upstream name servers, i.e., at least one of the 'send-client-subnet', 'client-subnet-zone' or 'client-subnet-always-forward' options is used. Resolvers supporting ECS need to segregate outgoing queries to accommodate for different outgoing ECS information. This re-opens up resolvers to a birthday paradox attack (Rebirthday Attack) that tries to match the DNS transaction ID in order to cache non-ECS poisonous replies.

FreshRSS is a self-hosted RSS feed aggregator. Prior to version 1.26.2, it's possible to poison feed favicons by adding a given URL as a feed with the proxy set to an attacker-controlled one and disabled SSL verifying. The favicon hash is computed by hashing the feed URL and the salt, whilst not including the following variables: proxy address, proxy protocol, and whether SSL should be verified. Therefore it's possible to poison a favicon of a given feed by simply intercepting the response of the feed, and changing the website URL to one where a threat actor controls the feed favicon. Feed favicons can be replaced for all users by anyone. Version 1.26.2 fixes the issue.

A vulnerability in client join services of Cisco Webex Meetings could allow an unauthenticated, remote attacker to manipulate cached HTTP responses within the meeting join service. This vulnerability is due to improper handling of malicious HTTP requests to the affected service. An attacker could exploit this vulnerability by manipulating stored HTTP responses within the service, also known as HTTP cache poisoning. A successful exploit could allow the attacker to cause the Webex Meetings service to return incorrect HTTP responses to clients.

An acceptance of extraneous untrusted data with trusted data vulnerability has been identified in Moxa’s Ethernet switches, which allows attackers with administrative privileges to manipulate HTTP Host headers by injecting a specially crafted Host header into HTTP requests sent to an affected device’s web service. This vulnerability is classified as Host Header Injection, where invalid Host headers can manipulate to redirect users, forge links, or phishing attacks. There is no impact to the confidentiality, integrity, and availability of the affected device; no loss of confidentiality, integrity, and availability within any subsequent systems.

check-jsonschema is a CLI and set of pre-commit hooks for jsonschema validation. The default cache strategy uses the basename of a remote schema as the name of the file in the cache, e.g. `https://example.org/schema.json` will be stored as `schema.json`. This naming allows for conflicts. If an attacker can get a user to run `check-jsonschema` against a malicious schema URL, e.g., `https://example.evil.org/schema.json`, they can insert their own schema into the cache and it will be picked up and used instead of the appropriate schema. Such a cache confusion attack could be used to allow data to pass validation which should have been rejected. This issue has been patched in version 0.30.0. All users are advised to upgrade. A few workarounds exist: 1. Users can use `--no-cache` to disable caching. 2. Users can use `--cache-filename` to select filenames for use in the cache, or to ensure that other usages do not overwrite the cached schema. (Note: this flag is being deprecated as part of the remediation effort.) 3. Users can explicitly download the schema before use as a local file, as in `curl -LOs https://example.org/schema.json; check-jsonschema --schemafile ./schema.json`

OpenClaw before 2026.3.24 contains an arbitrary code execution vulnerability in local plugin and hook installation that allows attackers to execute malicious code by crafting a .npmrc file with a git executable override. During npm install execution in the staged package directory, attackers can leverage git dependencies to trigger execution of arbitrary programs specified in the attacker-controlled .npmrc configuration file.

In JetBrains WebStorm before 2024.3 code execution in Untrusted Project mode was possible via type definitions installer script

A local user could edit the VideoEdge configuration file and interfere with VideoEdge operation.

Nuxt is an open-source web development framework for Vue.js. In Nuxt versions 3.1.0 to before 3.21.6 and 4.0.0-alpha.1 to before 4.4.6 and @nuxt/nitro-server versions 3.20.0 to before 3.21.6 and 4.0.0-alpha.1 to before 4.4.6, the /__nuxt_island/* endpoint accepts attacker-controlled props query/body parameters and renders any island component without verifying that the URL-resident hash (<Name>_<hashId>.json) was actually issued for those inputs by <NuxtIsland>. The hash is computed and embedded client-side but never validated server-side, so the same path can return materially different responses depending on the query. This issue has been patched in versions 3.21.6 and 4.4.6.

In JetBrains IntelliJ IDEA before 2025.3 missing confirmation allowed opening of untrusted remote projects over SSH