CVE-2026-103866 is a high-severity vulnerability in Apache Sling Engine before version 3.0.4. It involves inconsistent interpretation of HTTP requests, specifically HTTP request/response smuggling, due to the engine's query/form parser accepting non-standard percent-escapes and incorrect character sets while continuing request processing despite these errors. This can lead to request smuggling attacks. The issue is fixed in version 3.0.4.
CVE-2026-118516 is a vulnerability in the Symfony Routing component where the CompiledUrlMatcher and UrlMatcher interpret host and path information differently. This discrepancy allows path-based authorization rules to be bypassed 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 issue arises when routes define a host with placeholders that can consume parts of the path, causing unauthorized requests to reach protected controllers.
CVE-2026-120635 affects CGI versions before 5.00 for Perl and involves a multipart/form-data parsing vulnerability. The parser incorrectly matches boundaries anywhere instead of at the start of a line, mishandles boundaries for certain User-Agent values, and corrupts parts near the 4096-byte read size. This allows form fields to be smuggled past proxies and web application firewalls (WAFs).
@fastify/http-proxy versions 9.0.0 through 11.6.5 contain a path traversal vulnerability due to improper handling of URL-decoded route parameters in the rewritePrefix option. This allows unauthenticated remote attackers to bypass path restrictions and potentially cause server-side request forgery or authorization bypass in the proxied upstream service. The issue is fixed in version 11.6.6, and upgrading is the only mitigation.
CVE-2026-61789 is a vulnerability in bytebase's dbhub prior to version 0.22.6 that allows bypassing read-only SQL validation on MySQL and MariaDB connectors. The issue arises from a mismatch in comment parsing: the validator treats every double dash as a comment start, but MySQL and MariaDB require specific characters after the double dash. This discrepancy allows an attacker to hide write or destructive SQL statements from the validator while the database executes them. PostgreSQL is not affected. The vulnerability is fixed in version 0.22.6.
A vulnerability in Nuxt's nitro-server versions 4.6.0 through 4.6.2 allows improper resolution of path equivalence. The server resolves dot segments in URLs before rendering, but security checks evaluate the unresolved path, potentially bypassing route-based protections. This can lead to unauthorized rendering of pages like admin.vue when accessed via encoded paths such as /%2e/admin. The issue is fixed in version 4.6.3.
Forge (also called node-forge) is a native implementation of Transport Layer Security in JavaScript. Prior to 1.4.1, node-forge parsed deprecated extension OIDs 2.5.29.10 and 2.5.29.13 as basicConstraints, and pki.verifyCertificateChain() used the first extension with that name. Exploitation requires a CA trusted by the application to issue a certificate carrying an attacker-supplied deprecated-OID extension, commonly because the CA copies extensions from a certificate signing request. By placing a permissive deprecated extension before the genuine 2.5.29.19 basicConstraints extension, the certificate holder could have the certificate accepted as a CA and issue certificates for arbitrary names, enabling server or client impersonation. If the issuing CA adds a keyUsage extension without keyCertSign, node-forge rejects the chain. This issue is fixed in version 1.4.1.
CVE-2026-116272 is a vulnerability in MIME::Field::ParamVal versions before 5.521 for Perl's MIME-tools. It involves improper handling of duplicate boundary parameters when the parameter name contains nested comments. This causes the software to store the parameter under a literal name instead of the expected 'boundary', allowing a remote sender to hide an attachment from certain filters while the recipient's mail client still displays it.
CVE-2026-116269 is a vulnerability in MIME::Field::ParamVal versions before 5.521 for Perl. It causes the recommended_filename() method to return an empty string when processing RFC 2231 filenames with hyphenated language tags, such as 'utf-8'en-US'. This occurs because the decoder fails to properly strip the language suffix when it is not a single word, leading to an empty recommended filename instead of the actual filename. This behavior differs from other mail readers that correctly return the filename. As a result, a remote sender can bypass file extension checks on the recommended filename, potentially allowing disallowed attachments to pass through security policies.
CVE-2026-116266 is a vulnerability in MIME::Head versions from 5.518 up to but not including 5.521 for Perl. The multipart_boundary() method improperly decodes partly encoded multipart boundaries without marking the parse as ambiguous, allowing a remote sender to craft boundaries that evade filters relying on ambiguous_content(). This can cause attachments to be hidden from such filters while still being visible to the recipient's mail client.
CVE-2026-116265 is a vulnerability in MIME::Field::ParamVal versions before 5.521 for Perl's MIME-tools. It involves ambiguous multipart boundary interpretation due to incomplete checks on RFC 2231 continuation parameters. This ambiguity allows a remote sender to craft a message that splits differently on the recipient's mail client compared to other mail readers, potentially hiding attachments from filters relying on ambiguous_content().
CVE-2026-116264 is a vulnerability in MIME::Field::ParamVal versions before 5.521 for Perl where an invalid Content-Type parameter can cause subsequent parameters, including the multipart boundary, to be ignored. This leads to MIME::Parser treating the multipart body as a single undecoded part without marking the parse as ambiguous. As a result, a remote sender can hide an attachment from filters relying on ambiguous_content() while the recipient's mail client still displays it.
CVE-2026-116695 is a vulnerability in the Ruby CSS parser Crass prior to version 1.0.8. The parser incorrectly treats a closing parenthesis after an escaped backslash as escaped when processing malformed URLs, causing it to retain CSS declarations inside a bad URL token. This behavior can allow CSS-based UI redress, external resource loading, or data exfiltration via CSS, but does not enable script execution in modern browsers. The issue is fixed in Crass version 1.0.8.
@fastify/express versions 4.0.5 through 4.0.8 contain a vulnerability where enabling the non-default ignoreDuplicateSlashes option causes a mismatch between Fastify's routing and Express middleware path matching. This allows requests using absolute-form targets to bypass middleware such as authentication or authorization on scoped routes, potentially granting unauthorized access. The issue is fixed in version 4.0.9.
CVE-2026-115905 is a critical vulnerability in the zip-rs zip2 Rust library affecting versions prior to 9.0.5. The flaw arises from inconsistent handling of filenames between the Central Directory File Header and Local File Header in ZIP archives, allowing crafted archives to cause a streaming extractor to write attacker-controlled files potentially overwriting important files. This issue is resolved in version 9.0.5.
Interpretation Conflict vulnerability in hackney_url in benoitc hackney allows Server Side Request Forgery via a percent-encoded bracketed IPv6 literal or an IP literal with a trailing dot. A host such as %5B%3A%3Affff%3A127%2E0%2E0%2E1%5D (decoding to [::ffff:127.0.0.1]) or 127.0.0.1. is not an IP address to a caller's allowlist built on uri_string:parse/1 and inet:parse_address/1, while hackney connects to the address it denotes. The same applies to a Location target when redirects are followed.
The IP-literal gate that hackney_url:normalize/2 applies to a percent-decoded host, added to close the percent-encoded IP bypass of CVE-2026-47076, passes the decoded value to inet_parse:address/1 as is. The connect path in hackney_happy:parse_address/1 drops a surrounding bracket pair, and the resolver ignores a trailing dot, so both forms pass the gate and still reach the literal address. Domain names with a trailing dot and unencoded IPv6 literals are not affected.
This issue affects hackney: from 4.0.1 before 4.9.0.
A vulnerability in Nuxt versions from 3.11.0 up to but not including 3.21.4 and from 4.0.0 up to but not including 4.4.4 allows percent-encoded URL segments to bypass route rule checks. This causes appMiddleware to be skipped during server-side rendering, enabling unauthenticated users to access protected pages and their server-rendered data. The issue also affects app-manifest rules such as redirect, appLayout, ssr, prerender, and payload flags. The vulnerability is fixed in Nuxt versions 3.21.4 and 4.4.4.
Apache bRPC versions prior to 1.18.0 contain an interpretation conflict in HTTP path handling that allows remote attackers to bypass access control rules enforced by a front proxy. The issue arises because bRPC collapses consecutive slashes in request paths, treating paths with empty segments as equivalent to their collapsed form, contrary to RFC 3986. This discrepancy enables attackers to access resources blocked by the proxy by using paths with empty segments. Additionally, the HTTP/2 implementation does not properly validate the :path pseudo-header, allowing malformed requests to bypass checks. Version 1.18.0 addresses these issues by rejecting such malformed paths by default.
@fastify/middie versions 9.0.0 through 9.3.5 contain a vulnerability where path-scoped middleware may be bypassed due to a conflict in how mount paths are joined with plugin scope prefixes. This can cause middleware such as authentication or authorization to be silently skipped while the route handler still executes. The issue is fixed in version 9.3.6, and users should upgrade to this version or later. As a workaround, authorization can be enforced in a Fastify hook like preHandler.
fast-uri versions 2.4.2 through 2.4.7, 3.1.3 through 3.1.8, and 4.0.1 through 4.2.2 contain an improper input validation vulnerability in the canonicalizeHost() function. This function passes the parsed host to the WHATWG URL parser but does not verify that the entire input was consumed as a host, allowing silent rewriting of the authority component. Specifically, data before an at-sign is dropped and certain whitespace characters are stripped without error. This can cause host-based policies to be bypassed, potentially redirecting traffic to unintended destinations. The issue is fixed in versions 2.4.8, 3.1.9, and 4.2.3 and later.