CVE-2026-9734

NETGEAR R6250 before 1.0.4.6.Beta, R6400 before 1.0.1.18.Beta, R6700 before 1.0.1.14.Beta, R6900, R7000 before 1.0.7.6.Beta, R7100LG before 1.0.0.28.Beta, R7300DST before 1.0.0.46.Beta, R7900 before 1.0.1.8.Beta, R8000 before 1.0.3.26.Beta, D6220, D6400, D7000, and possibly other routers allow remote attackers to execute arbitrary commands via shell metacharacters in the path info to cgi-bin/.

Multiple cross-site request forgery (CSRF) vulnerabilities in D-Link DIR-600 router (rev. Bx) with firmware before 2.17b02 allow remote attackers to hijack the authentication of administrators for requests that (1) create an administrator account or (2) enable remote management via a crafted configuration module to hedwig.cgi, (3) activate new configuration settings via a SETCFG,SAVE,ACTIVATE action to pigwidgeon.cgi, or (4) send a ping via a ping action to diagnostic.php.

Multiple cross-site request forgery (CSRF) vulnerabilities in the HTTP Administration component in Cisco IOS 12.4 on the 871 Integrated Services Router allow remote attackers to execute arbitrary commands via (1) a certain "show privilege" command to the /level/15/exec/- URI, and (2) a certain "alias exec" command to the /level/15/exec/-/configure/http URI. NOTE: some of these details are obtained from third party information.

Network-AI, a TypeScript/Node.js multi-agent orchestrator, has a shipped, exported, documented feature called `ApprovalInbox` (`lib/approval-inbox.ts`). It is the network surface of the human-in-the-loop Approval Gate, which `ApprovalGate` uses to require explicit human approval for high-risk operations. The HTTP server it exposes has no authentication of any kind and sets `Access-Control-Allow-Origin: *` on every route, including the state-changing `POST /approvals/:id/approve` and `/deny`. As a result, in versions 5.0.0 through 5.12.1, any party who can send an HTTP request to the inbox port — a co-located process, a container/SSRF on the same host, a remote client when the operator binds a non-loopback address, or any website the operator visits in a browser (via the wildcard CORS) — can enumerate pending approvals and approve them, defeating the entire human-in-the-loop control and causing the gated high-risk action (e.g. a shell command the agent was holding for review) to execute without consent. This issue is fixed in v5.12.2. `ApprovalInbox` now accepts a `secret` option. When set, the mutating endpoints `POST /:id/approve` and `POST /:id/deny` require an `Authorization: Bearer <secret>` header, validated in constant time with `crypto.timingSafeEqual`. `startServer()` already binds to `127.0.0.1` by default; operators exposing the inbox on a network must set a secret.

A CSRF vulnerability exists in the `zone-include.php` script in Revive Adserver 6.0.7. Linking and unlinking banners or campaigns to zones could be triggered via crafted GET or POST requests without any verification of the CSRF token, allowing an attacker to perform these actions on behalf of an authenticated administrator.

dataCycle is a data management system for centrally storing, managing, searching, finding, and distributing data. In dataCycle-CORE, the module handling core processing and framework rules, before and including version 25.07.3, the application exposes server-side state changes through `GET` routes. Because browsers automatically send cookies on same-site top-level navigation and Rails does not apply CSRF protections to `GET`, an attacker can force a logged-in victim to modify application state by embedding a link, image, iframe, or redirect to one of these endpoints. This was confirmed on the target with a normal `Standard` account: a cross-site-style `GET` to `watch_lists/:id/add_item?thing_id=...` inserted content into a watch list with no CSRF token. Additional `GET` mutation routes exist in the codebase, including user impersonation for authorized admins and cache or translation state changes. This is patched in version 26.06.08.