CVE-2026-53754

CVE-2026-53754 affects Crawl4AI prior to version 0.8.8. The Docker API server’s SSRF protection (validate_webhook_url/validate_url_destination) used an explicit IPv4/IPv6 CIDR blocklist that missed several address families, allowing an unauthenticated attacker to reach internal services and cloud...

Cloudflare Quiche: Use-after-free in connection ID iterator FFI functions

Impact Cloudflare Quiche was affected by 2 use-after-free vulnerabilities in the connection ID iterator FFI functions. The quicheconnectioniditernext and quicheconnretiredscidnext functions would return a pointer to a ConnectionId to the applications via function arguments, but the the owned...

xorg-x11-server: xorg-x11-server-Xwayland: xorg-x11-server: out-of-bounds read/write in GLX ChangeDrawableAttributes

An out-of-bounds read flaw was found in the X.Org X server and Xwayland in glXDispChangeDrawableAttributes. A wrong size validation check can read a client-controlled number of bytes, exceeding the request buffer, leading to information disclosure. A write path also exists but requires byte-swapp...

Security update for cyrus-imapd (important)

openSUSE Security Update: Security update for cyrus-imapd Announcement ID: openSUSE-SU-2026:0204-1 Rating: important References: 1241536 1241543 1246165 1251788 Cross-References: CVE-2025-23394 CVE-2025-49812 CVSS scores: CVE-2025-49812 SUSE: 8.3...

CVE-2026-25559

OpenBullet2 through version 0.3.2 contains a path traversal vulnerability in the wordlist endpoint that allows authenticated attackers to perform arbitrary file read, write, and delete operations by supplying unsanitized absolute paths to the upload handler and wordlist functions. Attackers can...

CVE-2026-50262

An out-of-bounds read flaw was found in the X.Org X server and Xwayland in glXDispChangeDrawableAttributes. A wrong size validation check can read a client-controlled number of bytes, exceeding the request buffer, leading to information disclosure. A write path also exists but requires byte-swapp...

SkillGuard: A Permission Framework for Agent Skills

Agent skills extend LLM agents with reusable instructions, scripts, tool bindings, and contextual dependencies. However, current skill ecosystems largely rely on trust-based loading and static inspection, leaving a gap between what a skill can inject into an agent's context and what it can cause...

Astra Linux - уязвимость в linux, linux-5.10

A flaw was discovered in the Linux kernel’s implementation of Pressure Stall Information. Although this feature is disabled by default, it could allow an attacker to crash the system or cause other memory-corruption side effects...

The End of Trust: How Agentic AI Breaks Security Assumptions

For decades, the security of digital interaction has rested on an unacknowledged economic constraint. Attackers faced a tradeoff between the fidelity of a deception and the scale at which it could be deployed. Convincing impersonation required sustained human effort and was confined to a narrow s...

PT-2026-37186

Name of the Vulnerable Software and Affected Versions Heimdall versions prior to 0.17.14 Description Heimdall handles URL-encoded slashes %2F in a case-sensitive manner, whereas percent-encoding is defined as case-insensitive. When the allow encoded slashes variable is set to off the default...

CVE-2026-6043 Insecure Default Configuration in P4 Server

P4 Server versions prior to 2026.1 are configured with insecure default settings that, when exposed to untrusted networks, allow unauthenticated attackers to create arbitrary user accounts, enumerate existing users, authenticate to accounts with no password set, and access depot contents via the...

CVE-2026-6043 Insecure Default Configuration in P4 Server

P4 Server versions prior to 2026.1 are configured with insecure default settings that, when exposed to untrusted networks, allow unauthenticated attackers to create arbitrary user accounts, enumerate existing users, authenticate to accounts with no password set, and access depot contents via the...

EUVD-2026-25415

P4 Server versions prior to 2026.1 are configured with insecure default settings that, when exposed to untrusted networks, allow unauthenticated attackers to create arbitrary user accounts, enumerate existing users, authenticate to accounts with no password set, and access depot contents via the...

CVE-2026-33260

An attacker can send a web request that causes unlimited memory allocation in the internal web server, leading to a denial of service. The internal web server is disabled by default...

llm-security-lab

LLM Security Lab Laboratoire de sécurité pour application...

CVE-2026-33260

CVE-2026-33260 describes an input-validation flaw in the internal web server that can cause unlimited memory allocation when processing a web request, resulting in denial of service. The issue is documented across multiple feeds (NVD, ENISA EUVD, Debian OSV, CIRCL, etc.), all noting that the inte...

CVE-2026-33257 Insufficient input validation of internal webserver

An attacker can send a web request that causes unlimited memory allocation in the internal web server, leading to a denial of service. The internal web server is disabled by default...

CVE-2026-33257

The CVE-2026-33257 issue enables an attacker to send a web request that causes unlimited memory allocation in the internal web server, leading to a denial of service. The internal web server is disabled by default. In the provided documents, no concrete product/vendor/version, root cause details ...

CVE-2026-33257

An attacker can send a web request that causes unlimited memory allocation in the internal web server, leading to a denial of service. The internal web server is disabled by default...

PT-2026-34320

An attacker can send a web request that causes unlimited memory allocation in the internal web server, leading to a denial of service. The internal web server is disabled by default...