Vulnérabilités récentes · Page n° 172

decode-uri-component: Denial of service via quadratic replacement of percent-encoded runs

decode-uri-component provides a more robust alternative to decodeURIComponent. Prior to 0.5.1, decodeUriComponent in index.js falls back to customDecodeURIComponent after native decoding throws, and the fallback collects every distinct percent-encoded run before performing a full-string regular-expression replacement for each entry. Malformed input containing many distinct runs can therefore make both the number of replacements and the scanned input length grow together, causing quadratic CPU consumption and blocking the Node.js event loop; request-body inputs can reach the demonstrated size, while ordinary URL header limits reduce the practi

• Mettez à jour vers 0.5.1

@hono/node-server: `serveStatic` decodes the request path a second time, leading to bypass of middleware on static paths

@hono/node-server allows running the Hono application on Node.js. From 1.19.10 until 2.1.3, serveStatic decodes a request path a second time after Hono has already decoded it for routing, so a malformed percent-encoded sequence can make route middleware evaluate an encoded path while static file resolution uses a different directory. An unauthenticated attacker can bypass middleware mounted on a protected static prefix and read a file inside the configured static root, although the flaw does not permit traversal outside that root. This issue is fixed in version 2.1.3.

• Mettez à jour vers 2.1.3

Vite: `server.fs.deny` bypassed with queries

Vite is a frontend tooling framework for JavaScript. From 6.4.4 until 6.4.5, the Vite development server can evaluate server.fs.deny before stripping loader query parameters such as ?raw, ?import&raw, or ?import&url&inline. When the development server is exposed to the network and a sensitive file is inside server.fs.allow but matched by server.fs.deny, an unauthenticated attacker can append one of these queries and retrieve the denied file. This issue is fixed in version 6.4.5.

• Mettez à jour vers 6.4.5

Happy DOM: Cookie scope forgery: no DNS-label boundary on Domain= matching , attacker origin sets victim cookies

Happy DOM is a JavaScript implementation of a web browser without its graphical user interface. Prior to 20.14.6, Happy DOM's CookieContainer in src/cookie/CookieContainer.ts does not correctly validate or match Domain cookie attributes. It can accept a Domain unrelated to the origin setting the cookie, including a public suffix, and separately treats an unrelated suffix-sharing hostname as a match because cookieMatchesURL does not enforce a DNS-label boundary. In server-side rendering and prerendering applications that contact attacker-controlled origins and then make requests to a victim origin, these behaviors can plant and deliver a forge

• Mettez à jour vers 20.14.6

Happy DOM: SSRF via iframe and link stylesheet when parsing untrusted HTML

Happy DOM is a JavaScript implementation of a web browser without its graphical user interface. Prior to 21.0.0, Happy DOM loads iframe src URLs and link rel="stylesheet" href URLs by default while parsing untrusted HTML on the server because disableIframePageLoading and disableCSSFileLoading default to false. An attacker who controls parsed HTML can therefore cause server-side requests to arbitrary internal or cloud endpoints without an application opt-in and may expose confidential responses or credentials returned by those endpoints. This issue is fixed in version 21.0.0.

• Mettez à jour vers 21.0.0

Command Injection Vulnerability in Quantum Networks Router QN-I-470

This vulnerability exists in Quantum Networks router due to inadequate sanitization of user-supplied input in the management CLI interface. An authenticated remote attacker could exploit this vulnerability by injecting arbitrary OS commands on the targeted device. Successful exploitation of this vulnerability could allow the attacker to perform remote code execution with root privileges on the targeted device.

@happy-dom/server-renderer: ServerRendererServer Target-Origin Bypass SSRF / Open Proxy

Happy DOM is a JavaScript implementation of a web browser without its graphical user interface. Prior to 21.0.2, the Happy DOM server-renderer package's ServerRendererServer accepts unauthenticated HTTP/2 requests and resolves attacker-controlled :path values against server.targetOrigin in #onIncomingRequest in packages/@happy-dom/server-renderer/src/ServerRendererServer.ts without confirming that the resolved origin remains trusted before calling fetch. A protocol-relative path can change the destination to an arbitrary HTTP or HTTPS endpoint reachable by the server, including local and internal services, and the proxy relays the fetched res

Echo: Arbitrary Scheme Injection in WWWRedirect/NonWWWRedirect Middlewares via X-Forwarded-Proto

Echo is a Go web framework. Prior to 4.15.2 and 5.1.1, Context.Scheme returns values from X-Forwarded-Proto, X-Forwarded-Protocol, or X-Url-Scheme without validating that they are supported web schemes. WWWRedirect and NonWWWRedirect use that value to construct a Location header, so an attacker controlling a forwarded scheme header can cause a redirect to use an arbitrary URI scheme. Exploitation requires an affected redirect middleware and user interaction with the response; the HTTPS-specific redirect variants hardcode HTTPS and are unaffected. This issue is fixed in 4.15.2 and 5.1.1.

Echo: HTTPS Redirect and HSTS Bypass via Unconditional Trust of Forwarded Scheme Headers

Echo is a Go web framework. Prior to 4.16.0 and 5.4.0, Context.Scheme unconditionally trusts forwarded scheme headers from any client. A client that reaches Echo directly over plaintext HTTP can claim an HTTPS scheme and bypass HTTPSRedirect because the request is treated as already secure. Separately, behind a TLS-terminating proxy that passes an attacker-controlled header, a claimed HTTP scheme can suppress the Strict-Transport-Security header even though the browser receives the response over HTTPS. The practical risk is reduced when a trusted proxy overwrites the headers and direct backend access is prevented, but the framework previously

Echo: Open Redirect via Control-Character Prefix Defeating sanitizeURI

Echo is a Go web framework. Prior to 4.16.0 and 5.4.0, sanitizeURI checks only the first two bytes of a redirect target before deciding whether it begins with an absolute-path separator pair. When AddTrailingSlash or RemoveTrailingSlash is configured with a redirect status, a slash, TAB, separator path form can evade the guard, be emitted in the Location header, and be normalized by a browser into an attacker-controlled host. This issue is fixed in 4.16.0 and 5.4.0.

Cisco Secure Firewall Adaptive Security Appliance and Secure Firewall Threat Defense Software Remote Access SSL VPN Tunnel Denial of Service Vulnerability

A vulnerability in the Remote Access SSL VPN service of Cisco Secure Firewall Adaptive Security Appliance (ASA) Software and Cisco Secure Firewall Threat Defense (FTD) Software could allow an unauthenticated, remote attacker to cause the device to reload unexpectedly, resulting in a denial of service (DoS) condition. This vulnerability is due to logic errors when secondary authentication is configured on a tunnel group. An attacker could exploit this vulnerability by sending a crafted request when connecting to the SSL remote access VPN. A successful exploit could allow the attacker to cause the affected device to reload, resulting in a DoS c

Echo: Static Middleware Double-Unescape Path Traversal

Echo is a Go web framework. Prior to 4.15.3 and 5.2.0, middleware.Static can decode an already decoded Request.URL.Path a second time. Double-encoded separators can consequently be reconstructed before filesystem resolution. This can reach a file under a guarded route when the static middleware is registered globally. This issue is fixed in 4.15.3 and 5.2.0.

Ai-gateway-payload-processing: ai-gateway-payload-processing: tenant/credential selection from client-controlled url path with no in-process authorization

A flaw was found in ai-gateway-payload-processing. An authenticated remote attacker can exploit this vulnerability by submitting a request specifying another tenant's model identifier in the request body or header. Because the service does not verify authorization before retrieving credentials from the shared store, it injects the target tenant's application programming interface (API) keys into the request. This flaw allows an attacker to perform unauthorized requests and consume another tenant's large language model (LLM) resources and quotas.

Echo static file middleware route protection bypass via encoded dot-dot traversal

Echo is a Go web framework. Prior to 4.16.0 and 5.4.0, Echo's static file handlers can normalize dot segments after route matching and serve a file through a public static route even though the corresponding direct path is protected by route-level middleware. An unauthenticated attacker can read files that an application intended to protect with route-level middleware, when those files are also reachable through a static filesystem mount after path normalization. This vulnerability is the result of an incomplete fix for CVE-2026-55677. This issue is fixed in 4.16.0 and 5.4.0.

Kserve: kserve: server-side request forgery via malicious storageuri

A flaw was found in KServe. A namespace-scoped user with create/update permissions on InferenceService custom resources can exploit this vulnerability by providing a malicious `storageUri`. This Server-Side Request Forgery (SSRF) vulnerability allows the `storage-initializer` init container to issue arbitrary outbound HTTP/HTTPS requests from inside the cluster. This can lead to information disclosure, including credential theft for the node's Identity and Access Management (IAM) role or access to internal cluster services.

mcp-sanitizer: URL sanitizer allows localhost/IPv6 loopback SSRF bypass via explicit port or bracketed IPv6

## URL sanitizer allows localhost/IPv6 loopback SSRF bypass via explicit port or bracketed IPv6 ### Summary `mcp-sanitizer` v1.3.3 fails to block Server-Side Request Forgery (SSRF) targets when a `localhost` URL includes an explicit port (e.g., `http://localhost:8080/`) or when an IPv6 loopback/link-local address is expressed in bracketed notation (e.g., `http://[::1]:8080/`, `http://[fe80::1]:8080/`). Because the library is explicitly designed to gate MCP tool parameters such as `web_fetch` and `web_scraper` before outbound HTTP requests are made, a bypass here directly enables an attacker-controlled URL to reach internal services, leaking

• Mettez à jour vers 1.4.0

mcp-sanitizer: File Path Sanitizer Permits Arbitrary User-Home Absolute Paths

## File Path Sanitizer Permits Arbitrary User-Home Absolute Paths ### Summary `mcp-sanitizer` v1.3.3 explicitly whitelists `/home/` and `/Users/` as allowed absolute path prefixes in its file path validation logic (`src/utils/validation-utils.js:188-196`). As a result, paths such as `/home/alice/.ssh/id_rsa` and `/home/alice/.aws/credentials` pass through `sanitizer.sanitize()` with `blocked=false`, even though they point to highly sensitive files outside any configured safe root. Any consumer application that trusts the sanitized path and passes it to a filesystem read/write operation (e.g., `fs.readFile`) will expose arbitrary user-home d

• Mettez à jour vers 1.4.0

rucio has an Authorization Bypass on REST API endpoint for signing URLs

## Summary The `/credentials/signurl` endpoint decides authorization from the client-supplied `X-Rucio-Account` HTTP header instead of the authenticated token's account. Any client holding **any** valid auth token (even a low-privilege account they legitimately own) can send `X-Rucio-Account: root` to pass the root-only permission check and have the Rucio server mint a presigned **read / write / delete** URL for an **arbitrary** object, signed with the RSE's stored S3 / GCS / Swift credentials. ## Details `SignURL.get()` takes the authorization principal from a request header and never uses the token-authenticated identity: ```python # lib/

• Mettez à jour vers 35.9.0
• Mettez à jour vers 38.6.0
• Mettez à jour vers 41.0.0rc1

Gearboy advisory / Gearsystem body: Unauthenticated network exposure of MCP HTTP tools

Gearboy is a Game Boy, Game Boy Color, and Super Game Boy emulator and debugger with an embedded MCP server. Gearsystem, the affected component, has an optional `--mcp-http` mode that accepts unauthenticated MCP requests from any reachable network peer despite localhost-only documentation and logging. In the described `platforms/shared/desktop/mcp/mcp_transport.h` transport, the socket binds `INADDR_ANY`, normally on port 7777, and POST requests to `/mcp` reach dispatch without `Authorization`, `Host`, or `Origin` checks. Responses and OPTIONS advertise `Access-Control-Allow-Origin: *`. A peer can initialize an MCP session and invoke exposed

Ember: Unneutralised terminal escape/control sequences from Caddy logs injected into the operator's TUI

Ember is a real-time terminal dashboard for Caddy servers with FrankenPHP support. Prior to 1.4.2, interactive TUI row formatters pass Caddy access-log fields, particularly the request URI, through the `fitCellLeft` render helper without neutralizing terminal escape or control sequences. An unauthenticated remote client can place ANSI, OSC, or CSI control bytes in an ordinary HTTP request target. Caddy JSON-escapes those bytes, `ParseLogLine` decodes them back into raw controls, and the formatters concatenate them into `View()` output reaching `os.Stdout`. When the operator views the default `Logs` tab, the terminal interprets these sequences