Exploited CVEs
Vulnerabilities with confirmed exploitation — sourced from threat intelligence reports with associated IoCs and actor attribution.
Race condition in tempfile.TemporaryDirectory cleanup allows deleting files outside the temporary directory
Out-of-Bounds Write in Google Brotli Decoder
Pillow: Heap buffer overflow in ImagingJpegDecode
Pillow: Heap buffer overflow in _tiffWriteProc when saving TIFF to BytesIO
Pillow: Argument Injection in Ghostscript call, potentially leading into RCE
LiteLLM: SQL injection in Proxy API key verification
Drupal core - Highly critical - SQL injection - SA-CORE-2026-004
OS command injection in mcp-remote when connecting to untrusted MCP servers
Integer Overflow or Wraparound in GitLab
Improper Limitation of a Pathname to a Restricted Directory ('Path Traversal') in GitLab
A vulnerability exists in NGINX Plus and NGINX Open Source when a map directive uses regex matching and a string expression references the map's regex capture variables before referencing the map output variable. Alternatively, the same result could be achieved by using a non-cacheable variable in a string expression under certain conditions. An unauthenticated attacker along with conditions beyond their control can exploit this vulnerability by sending crafted HTTP requests. This may cause a he
Samsung QA-Board — Unauthenticated OS command injection via output redo/tuning
Collabora Online Built-in CODE Server before 26.4.303 (richdocumentscode), a Nextcloud app, allows remote code execution, as exploited in the wild in September and October 2026, a different vulnerability than CVE-2025-66208. This is an OS command injection in proxy.php when the CODE server isn't running (e.g., the product was recently restarted). The vulnerability is not present in standalone Collabora Online servers; only this deployment with CODE Server is affected.
Langflow: OS command injection (RCE) via arbitrary command in MCP stdio server configuration
Flowise: CSV Agent Remote Code Execution via Pyodide Code Injection — Root Shell Verified
Langflow: Authenticated RCE via MCP Stdio transport allows any user to execute arbitrary OS commands on the server
vhost: reset the vring metadata cache on vring reconfiguration
A heap-based buffer overflow vulnerability exists in the lossless_jpeg_load_raw functionality of LibRaw Commit 0b56545 and Commit d20315b. A specially crafted malicious file can lead to a heap buffer overflow. An attacker can provide a malicious file to trigger this vulnerability.
Ghostscript < 10.08.0 Heap Buffer Overflow via JPEG 2000 Output Adapter
Improper Neutralization of Special Elements Used in a Template Engine in GitLab AI Gateway
The WPGraphQL 0.2.3 plugin for WordPress allows remote attackers to register a new user with admin privileges, whenever new user registrations are allowed. This is related to the registerUser mutation.
An issue has been discovered in GitLab CE/EE affecting all versions starting from 11.9. GitLab was not properly validating image files that were passed to a file parser which resulted in a remote command execution.
named in ISC BIND 9.x before 9.9.7-P2 and 9.10.x before 9.10.2-P3 allows remote attackers to cause a denial of service (REQUIRE assertion failure and daemon exit) via TKEY queries.
GNU Bash through 4.3 bash43-026 does not properly parse function definitions in the values of environment variables, which allows remote attackers to execute arbitrary commands via a crafted environment, as demonstrated by vectors involving the ForceCommand feature in OpenSSH sshd, the mod_cgi and mod_cgid modules in the Apache HTTP Server, scripts executed by unspecified DHCP clients, and other situations in which setting the environment occurs across a privilege boundary from Bash execution.
Directory traversal with remote code execution can occur in /upload in ONLYOFFICE Document Server before 5.6.3, when JWT is used, via a /.. sequence in an image upload parameter.
Strapi through 4.5.5 allows attackers (with access to the admin panel) to discover sensitive user details by exploiting the query filter. The attacker can filter users by columns that contain sensitive information and infer a value from API responses. If the attacker has super admin access, then this can be exploited to discover the password hash and password reset token of all users. If the attacker has admin panel access to an account with permission to access the username and email of API use
Apache Struts 2.3.19 to 2.3.20.2, 2.3.21 to 2.3.24.1, and 2.3.25 to 2.3.28, when Dynamic Method Invocation is enabled, allow remote attackers to execute arbitrary code via method: prefix, related to chained expressions.
The mod_copy module in ProFTPD 1.3.5 allows remote attackers to read and write to arbitrary files via the site cpfr and site cpto commands.
An out-of-bounds write issue was addressed with improved checks to prevent unauthorized actions. This issue is fixed in visionOS 2.3.2, iOS 18.3.2 and iPadOS 18.3.2, macOS Sequoia 15.3.2, Safari 18.3.1, watchOS 11.4, iPadOS 17.7.6, iOS 16.7.11 and iPadOS 16.7.11, iOS 15.8.4 and iPadOS 15.8.4. Maliciously crafted web content may be able to break out of Web Content sandbox. This is a supplementary fix for an attack that was blocked in iOS 17.2. (Apple is aware of a report that this issue may have
vLLM: Loose `cache_salt` validation lets a single request kill EngineCore on LMCache-MP deployments — uncaught downstream `ValueError` denial of service