Missing authorization in Azure DNS allows an unauthorized attacker to elevate privileges over a network. CVSSv3.1 10.0 (CRITICAL)
Improper authentication in Microsoft Exchange Online allows an unauthorized attacker to perform tampering over a network. CVSSv3.1 10.0 (CRITICAL)
Server-side request forgery (ssrf) in Azure AI Search allows an authorized attacker to elevate privileges over a network. CVSSv3.1 8.5 (HIGH)
Heap-based buffer overflow in Microsoft Account allows an unauthorized attacker to execute code over a network. CVSSv3.1 9.8 (CRITICAL)
Improper authorization in Azure Red Hat OpenShift (ARO) allows an authorized attacker to elevate privileges over a network. CVSSv3.1 9.1 (CRITICAL)
Improper input validation in Microsoft Surface allows an authorized attacker to execute code over a network. CVSSv3.1 9.9 (CRITICAL)
Deserialization of untrusted data in M365 Copilot allows an authorized attacker to execute code over a network. CVSSv3.1 9.9 (CRITICAL)
Improper access control in Azure API Management (APIM) allows an authorized attacker to execute code over a network. CVSSv3.1 8.0 (HIGH)
Trend Micro documents JADEPUFFER, the first autonomous ransomware operation where an LLM agent executed a complete intrusion chain—from initial access via unpatched Langflow RCE (CVE-2025-3248) through credential harvesting, privilege escalation via Nacos auth bypass (CVE-2021-29441), persistence, and database encryption—without human operator intervention. The agent adapted in real-time to obstacles, rewrote its own code on failure, and generated 600+ distinct payloads, though the monetization layer failed due to unsaved encryption keys and a placeholder ransom address. The attack shifts defensive burden from indicator-based detection to behavior-based monitoring, as agent-run attacks produce disposable per-victim indicators and improvised infrastructure.
Elastic Security Labs published a detailed technical analysis of agentic SOC architectures, comparing specialized multi-agent workflows versus single agents with on-demand skills across 36,822 production conversations. The specialized workflow achieved 5.7× cost reduction ($0.69 vs $3.42 per investigation) through inlined methodology and deterministic routing, while the single-agent approach excels for interactive analyst-led investigations.
Pronetiqs IntraVUE versions 3.2.1a14 and prior have an unintended proxy or intermediary vulnerability which could allow an attacker to use an active proxy, which would bypass OT segmentation. CVSSv3.1 10.0 (CRITICAL)
Pronetiqs IntraVUE versions 3.2.1a14 and prior have an exposure of sensitive system information to an unauthorized control sphere vulnerability which could expose the underlying host/share filesystem. CVSSv3.1 8.6 (HIGH)
Skipper contains an incomplete fix for CVE-2026-50197 in which oversized request bodies bypass Open Policy Agent (OPA) deny-on-presence Rego policies. When a request body exceeds the configured maxBodyBytes limit, Skipper forwards the full payload to the upstream service while OPA evaluates against an empty parsed_body, so policies that deny requests based on body content are not enforced and forbidden actions proceed. No fixed version is available; v0.27.26 adds documentatio CVSSv3.1 8.2 (HIGH)
9router 0.4.59 (fixed in 0.4.60) contains a chain of vulnerabilities: a hardcoded default password (123456) that authenticates any fresh installation, a bypass of the LOCAL_ONLY network gate via a spoofed Host header, and unvalidated arguments passed to child_process.spawn() when registering MCP plugins. A remote, unauthenticated attacker can log in with the default credential, spoof the Host header to reach local-only routes, and register a malicious MCP plugin (e.g. node -e CVSSv3.1 9.9 (CRITICAL)
Out of bounds write in Codecs in Google Chrome prior to 150.0.7871.186 allowed a remote attacker to potentially perform a sandbox escape via a crafted HTML page. (Chromium security severity: High) CVSSv3.1 8.8 (HIGH)
Use after free in WebMCP in Google Chrome prior to 150.0.7871.186 allowed a remote attacker to execute arbitrary code inside a sandbox via a crafted HTML page. (Chromium security severity: High) CVSSv3.1 8.8 (HIGH)
Use after free in Blink in Google Chrome prior to 150.0.7871.186 allowed a remote attacker to execute arbitrary code inside a sandbox via a crafted HTML page. (Chromium security severity: High) CVSSv3.1 8.8 (HIGH)
Use after free in Input in Google Chrome prior to 150.0.7871.186 allowed a remote attacker who had compromised the renderer process to potentially perform a sandbox escape via a crafted HTML page. (Chromium security severity: High) CVSSv3.1 8.3 (HIGH)
Cal.com (calcom/cal.diy) before 5.9.9 is vulnerable to unauthenticated remote code execution because it bundles a version of Next.js whose React Server Components (RSC) request handling deserializes attacker-controlled input. A remote attacker can send a crafted RSC request to the server and cause arbitrary code to be executed during server-side processing, without authentication or user interaction. The flaw derives from the upstream Next.js vulnerability CVE-2025-55182 and CVSSv3.1 10.0 (CRITICAL)
Cal.com (calcom/cal.diy) versions through 4.7.15 contain a stored cross-site scripting vulnerability. The single booking view (e.g., https://app.cal.com/booking/<id>) renders booking-question field labels via React's dangerouslySetInnerHTML without sanitizing or escaping user input. An attacker who can create an event type with a malicious booking-question label can inject arbitrary HTML/JavaScript that executes when a victim opens the crafted booking URL. The issue is fixed CVSSv3.1 8.9 (HIGH)
cal.com (calcom repository, later renamed cal.diy) is affected by a repository takeover vulnerability in its GitHub Actions workflows. The workflow pr.yml uses the pull_request_target trigger with the repository's default write permissions and passes them down to check-types.yml. check-types.yml then performs a 'dangerous' checkout of the attacker-submitted pull request code (via the dangerous-git-checkout action) and subsequently executes it (through yarn install and package CVSSv3.1 9.9 (CRITICAL)
Cal.com (repository calcom/cal.diy) in versions <= 4.7.15 is vulnerable to cross-site scripting (XSS) on the publicly accessible single booking view (e.g., /booking/<id>). Booking question (form field) labels are rendered via React's dangerouslySetInnerHTML without proper input sanitization or CSP, so an attacker who can create an event type with a malicious booking question label can inject arbitrary HTML/JavaScript that executes when a victim visits the booking view URL. Se CVSSv3.1 8.9 (HIGH)
An issue in xiandafu beetl 3.20.2 allows a remote attacker to execute arbitrary code via the type.new function and the property reflection mechanism CVSSv3.1 9.8 (CRITICAL)
The affected product is vulnerable to a heap-based buffer overflow via a crafted MMS Initiate request. Remote code execution (RCE) has been demonstrated when ASLR is disabled; memory corruption or denial of service may occur in configurations where ASLR is enabled. CVSSv3.1 8.1 (HIGH)
nebula-mesh is a self-hosted control plane for Slack Nebula mesh virtual private network. Prior to version 0.3.4, the `/api/v1/*` route surface trusts the bearer token alone for authorisation on most endpoints. The codebase itself admits this at `internal/api/hosts.go:384`: "API trusts the bearer token for authorisation; per-CA ownership is enforced only in the Web layer." The Web UI gates state-changing routes through `loadAccessibleCA` (`internal/web/cas.go`); CA-management CVSSv3.1 9.9 (CRITICAL)