Local Privilege Escalation via Heap Overflow in Android hwcrypto_proxy

In handle_fips_hmac_req of hwcrypto_proxy.c, there is a possible heap buffer overflow due to missing bounds check. This could lead to local escalation of privilege with System execution privileges needed. User interaction is not needed for exploitation.

Android efw_ap_transport Local Privilege Escalation Non‑Heap Buffer Overflow

In RecordApTx of efw_ap_transport.c, there is a possible Non-Heap Buffer Overflow due to a buffer overflow. This could lead to local escalation of privilege with execution privileges needed. User interaction is for exploitation.

SvelteKit: Prerendered paths are interpreted as regular expressions in Vercel route config generated by @sveltejs/adapter-vercel and @sveltejs/adapter-static

SvelteKit is a framework for rapidly developing robust, performant web applications using Svelte. Prior to @sveltejs/adapter-vercel 6.3.5 and 7.0.2 and @sveltejs/adapter-static 3.0.11 and 4.0.1, an attacker-influenced link crawled during prerendering could be emitted into the generated Vercel routing configuration without escaping regular-expression metacharacters. The configuration could consequently interpret the path as regular-expression syntax, match unintended application paths, and expose content or behavior outside the intended path. This issue is fixed in @sveltejs/adapter-vercel 6.3.5 and 7.0.2 and @sveltejs/adapter-static 3.0.11 and 4.0.1.

Nuxt: Colliding island props can share one `/__nuxt_island` path (incomplete fix for CVE-2026-46342)

Nuxt is an open-source web development framework for Vue.js. From 3.21.6 to before 4.0.0 and from 4.4.6 until 4.6.4, the /__nuxt_island path hash uses ohash serialization that does not escape quotes in string values or delimiters in object keys, allowing different island-prop tuples to produce the same path. For prerendering, substituted content is written into the generated page only if the crawler reaches a colliding URL before rendering the page with the legitimate props. Separately, a CDN or reverse proxy that keys island responses by path alone can be primed with the colliding response. If the substituted prop reaches v-html, innerHTML, or another unsafe HTML sink, the response can execute script in the application origin; otherwise the impact is content substitution. Applications that do not use server components or NuxtIsland, islands without attacker-influenced string props, and responses not cached by path alone are unaffected. This issue is fixed in version 4.6.4.

In multiple locations, there is a possible way to zero arbitrary secure memory via unvalidated image body addresses due to a missing bounds check. This could lead to physical escalation of privilege with System execution privileges needed. User interaction is not needed for exploitation.

TeamSpeak 3 Server ECC Key heap-based overflow

A security vulnerability has been detected in TeamSpeak 3 Server up to 3.13.8. This vulnerability affects unknown code of the component ECC Key Parser. Such manipulation leads to heap-based buffer overflow. The attack may be launched remotely. Upgrading to version 3.13.9 is able to resolve this issue. It is suggested to upgrade the affected component. The vendor explains: "We have just released TeamSpeak 3 Server 3.13.9 to address a remaining variant of CVE-2026-4391, affecting ECC key parsing during the connection handshake. The original fix released with 3.13.8 did not cover this additional variant, which could allow an unauthenticated remote attacker to crash the server using specially crafted input."

A improper encoding or escaping of output vulnerability in Fortinet FortiAnalyzer 8.0.0 through 8.0.2, FortiAnalyzer 7.6.0 through 7.6.7, FortiAnalyzer Cloud 8.0.1 through 8.0.2, FortiAnalyzer Cloud 7.6.0 through 7.6.7, FortiAnalyzer-BigData 7.6.0 through 7.6.2, FortiManager 8.0.0 through 8.0.2, FortiManager 7.6.0 through 7.6.7, FortiManager Cloud 8.0.1 through 8.0.2, FortiManager Cloud 7.6.0 through 7.6.7 may allow an unauthenticated attacker to execute unauthorized code or commands via a specially crafted HTTP request

Cudy TR3000 before 2.5.42 Root RCE via OpenVPN Key Upload Injection

Cudy TR3000 before 2.5.42 contains a remote code execution vulnerability that allows authenticated administrators to execute commands as root by uploading crafted OpenVPN private or tls-auth keys. Attackers can close the inline key block with </key>, inject directives such as tls-verify that bypass the .ovpn denylist, and have OpenVPN execute the hook as root.

In multiple functions of AndroidManifest.xml, there is a possible way to trigger unauthorized firmware updates via exported broadcast receivers due to a missing permission check. This could lead to local escalation of privilege with no additional execution privileges needed. User interaction is not needed for exploitation.

In PresenceDecoderV1::DecodeDataElements of presence_decoder_v1.cc, there is a possible out-of-bounds write due to a missing bounds check. This could lead to remote escalation of privilege with no additional execution privileges needed. User interaction is not needed for exploitation.

In multiple functions of cn_Tai.cpp, there is a possible way to trigger an out of bounds write due to memory corruption. This could lead to remote (proximal/adjacent) code execution with no additional execution privileges needed. User interaction is not needed for exploitation.

In sub_layer_hrd_parameters of hevc_decoder, there is a possible Non-Heap Buffer Overflow due to a buffer overflow. This could lead to local escalation of privilege with execution privileges needed. User interaction is for exploitation.

Out-of-bounds write in Android NFC HAL leading to local privilege escalation

In StNfc_hal_write of hal_st21nfc.cc, there is a possible way to trigger an out of bounds write into global memory due to an incorrect bounds check. This could lead to local escalation of privilege with no additional execution privileges needed. User interaction is not needed for exploitation.

In knee_function_info of hevc_decoder, there is a possible Non-Heap Buffer Overflow due to a buffer overflow. This could lead to local escalation of privilege with execution privileges needed. User interaction is for exploitation.

In slice_header of hevc_decoder, there is a possible Non-Heap Buffer Overflow due to a buffer overflow. This could lead to local escalation of privilege with execution privileges needed. User interaction is for exploitation.

In acquireMemory of bt_bqr_offload.cc, there is a possible way to trigger an out of bounds write via a buffer size underflow due to an incorrect bounds check. This could lead to local escalation of privilege with no additional execution privileges needed. User interaction is not needed for exploitation.

Nokogiri: Mutation XSS through `noscript` content in Nokogiri's HTML5 serializer

Nokogiri is an open source XML and HTML library for the Ruby programming language. From 1.12.0 until 1.19.5, the C extension on CRuby and TruffleRuby has a Nokogiri::HTML5 serializer that does not escape text inside noscript elements or angle brackets inside attribute values. When an application sanitizes untrusted HTML with the default parsing options, allows noscript, and sends serialized output to a scripting-enabled browser, the browser can parse markup that Nokogiri did not produce, enabling mutation cross-site scripting. Using parse_noscript_content_as_text: true is the recommended workaround, not the vulnerable default. JRuby is not affected because it does not implement Nokogiri::HTML5. This issue is fixed in version 1.19.5.

The proxy-cache and proxy-cache-advanced plugins in Kong Gateway Enterprise and Kong AI Gateway do not escape separator characters when building the cache key, allowing a remote attacker to cause a response cached for one request to be served for a different request on the same Route.

Stored XSS in the @waline/client UserList widget via an unescaped commenter link field

Waline is a comment system with backend support. In versions 3.15.2 and prior, the @waline/client UserList widget builds its DOM by concatenating server-returned commenter fields directly into root.innerHTML with no HTML escaping. The commenter nick and link are stored and returned verbatim by the server (only the comment body is sanitized), and the link field is never sent to the spam filter. So an unauthenticated visitor who posts a comment with link = x"><img src=x onerror=...> achieves stored cross-site scripting that executes, with no interaction, in the browser of every visitor to a page that embeds the UserList widget. At time of publication, there are no publicly available patches.

Stored XSS in @waline/client via the unsanitized markdown preview of a comment's raw body on edit

Waline is a comment system with backend support. In versions 3.15.2 and prior, @waline/client renders the comment editor's live preview with marked and no sanitizer (utils/markdown.ts), and the preview is injected with v-html. When a comment is edited, the editor is seeded with the comment's raw, un-sanitized body (orig). Only administrators and the comment owner can edit. So an administrator who clicks Edit on an attacker's comment renders that attacker's raw HTML in their own session: an unauthenticated commenter achieves privilege-escalation stored XSS against a moderator with a single Edit click. At time of publication, there are no publicly available patches.