Newly released CVEs across every platform — sleek to read, verbose on demand, and served raw as markdown for AI and agent ingestion. Severity reads as depth: the deeper the contact, the graver the threat.
Depth = severity + exploitation
CVE-2026-116497Medium· 5.5A NULL pointer dereference vulnerability was found in gssproxy. When a GSSX_INIT_SEC_CONTEXT request is received with the target_name field absent, the daemon dereferences the resulting NULL pointer and crashes. The request requires no a…
CVE-2026-92171Medium· 6.5A flaw was found in the Infinispan Operator for Red Hat Data Grid. When cross-site (xsite) TLS is enabled, the operator launches the Gossip Router container with keystore and truststore passwords passed as plain-text command-line argumen…
CVE-2026-116124High· 7.8A flaw was found in libgsf. A heap-based buffer overflow vulnerability exists in the OLE2 document parser within the ole_info_read_metabat function. A remote attacker could exploit this flaw by enticing a user to open a specially crafted f…
CVE-2026-116137Low· 3.7A flaw was found in adm-zip. A remote attacker can exploit this issue by providing a specially crafted ZIP archive containing duplicate entries with equivalent path names to bypass validation checks. Consequently, an application may inspec…
CVE-2026-116141Medium· 5.3A flaw was found in ImageMagick. A remote attacker could cause a Denial of Service (DoS) through excessive resource consumption by submitting a specially crafted Ultra High Dynamic Range (UHDR) image file. Because the UHDR decoder validate…
CVE-2026-116140Low· 3.3A flaw was found in ImageMagick. An attacker could bypass administrator-defined security policies by convincing the application to process specially crafted input that triggers a delegate invocation. This issue occurs due to incorrect pars…
CVE-2026-116149High· 7.8A flaw was found in ImageMagick. A stack-based buffer overflow vulnerability exists in the handling of bzip2 compressed images, where stack memory can be overwritten during decoding. By submitting a specially crafted image, an attacker can…
CVE-2026-116148Low· 3.6A flaw was found in ImageMagick. The Magick Vector Graphics (MVG) encoder does not properly neutralize user-specified drawing options before writing them into generated output. An attacker capable of supplying drawing options can exploit t…
CVE-2026-116152Medium· 5.3A flaw was found in ImageMagick. An attacker could exploit this issue by providing a specially crafted MAT image file that triggers an error-handling path where allocated memory is not properly released. Over time, repeated processing of t…
CVE-2026-116153Low· 2.5A flaw was found in ImageMagick. An attacker could exploit this vulnerability to cause a Denial of Service (DoS) or potentially execute arbitrary code on 32-bit systems. By convincing a user or automated service to process a specially craf…
CVE-2026-116160Low· 3.3A flaw was found in ImageMagick. An attacker can cause a Denial of Service (DoS) by tricking a user into processing a specially crafted image file using the X11 animate feature. Processing the image triggers a heap-based buffer overflow, w…
CVE-2026-116172Medium· 5.9A flaw was found in vLLM. Due to improper input validation in the Rust gRPC interface, a remote attacker can bypass configured limits by sending requests with excessively large or wrapped token count values. When processed, these invalid v…
CVE-2026-116174Medium· 5.9A flaw was found in vLLM. A remote attacker can trigger a Denial of Service (DoS) by submitting requests containing out-of-bounds token identifiers to the gRPC service. Because the application fails to validate pre-tokenized input against …
CVE-2026-116177High· 7.5A flaw was found in vLLM. This flaw allows a remote, unauthenticated attacker to cause a Denial of Service (DoS) by exhausting system memory. The vulnerability occurs because the Nemotron-VL processor modules disable the global maximum ima…
CVE-2026-116176High· 8.2A flaw was found in vLLM. This vulnerability allows a remote attacker to trigger a Server-Side Request Forgery (SSRF) by submitting an HTTP request containing a crafted X-Request-Id header. Because the application extracts network addresse…
CVE-2026-116181Medium· 6.5A flaw was found in vllm. An authenticated remote attacker could cause a Denial of Service (DoS) or memory corruption by submitting requests containing invalid stop token identifiers. Because the application does not verify whether provide…
CVE-2026-116192Medium· 5.9A flaw was found in rclone. This vulnerability allows an unauthenticated remote attacker to cause information disclosure through Server-Side Request Forgery (SSRF). When the rc-serve feature is enabled, an attacker can provide a double-sla…
CVE-2026-116190Low· 3.7A flaw was found in rclone. When interacting with WebDAV storage services, rclone inadvertently forwards user-configured custom credential headers to external destinations across HTTP redirects. A remote attacker who controls or influences…
CVE-2026-116196Medium· 5.3A flaw was found in Rclone. This vulnerability allows information disclosure due to improper handling of HTTP redirects within the Backblaze B2 storage backend. When download requests are redirected, the client inadvertently forwards sensi…
CVE-2026-116194Medium· 6.8A flaw was found in rclone. A remote attacker could redirect file transfers and expose sensitive data by exploiting a Cross-Site Request Forgery (CSRF) vulnerability in the pCloud storage backend. During authorization, the local authentica…
CVE-2026-116189High· 7.5A flaw was found in rclone. A remote attacker with the ability to tamper with the release distribution path can bypass cryptographic signature verification during the self-update process. By prepending an unsigned checksum before the signe…
CVE-2026-115877Medium· 5.4A flaw was found in Twisted. The HTTP server component improperly accepts bare carriage return or line feed characters within header values instead of enforcing strict line endings. A remote attacker can exploit this framing disagreement a…
CVE-2026-106071Low· 3.6PoCA flaw was found in the Node.js Permission Model. Wildcard --allow-fs-read or --allow-fs-write rules can grant access to paths that do not match the intended wildcard when the requested path is shorter than the wildcard prefix and the radi…
CVE-2026-106072Medium· 6.7PoCA flaw was found in Node.js TLS hostname verification. When a client connects using a non-canonical IPv4 hostname spelling (for example 0177.0.0.1, 0x7f.0.0.1, 2130706433, or 127.1), tls.checkServerIdentity() may fail to classify the host …
CVE-2026-106073Low· 3.3PoCA flaw was found in the Node.js Permission Model. fs.realpathSync() does not perform the filesystem read permission checks required of comparable fs APIs. This is an incomplete fix of CVE-2026-21715, which only covered fs.realpathSync.nati…
The beta is over. VulnSea now runs at vulnsea.com, every old beta link redirects to the same page, and your account, API keys and watchlists came with it. Plus a round of hardening, faster pages, and new request limits agents should know about.
A summary of everything that shipped over the last two weeks — the whole corpus is open, agents get change feeds, alias resolution and EPSS movers, and the data now includes CVE.org, vendor CSAF, aggregated exploits and per-source scores.
A step-by-step guide to plugging VulnSea into automated and agentic workflows — poll the delta, triage without burning tokens, match an SBOM, and let an MCP-native model do the reasoning.