Vulnerabilities

Tesseract OCR 5.5.3 and earlier fails to validate counts read from the legacy INTTEMP component of .traineddata files: Classify::ReadIntTemplates uses NumClassPruners, NumClasses, and NumProtoSets as loop bounds without capping them against their maximums, and writes heap pointers past the fixed-capacity ClassPruners and ProtoSets arrays. An attacker who can supply or replace a .traineddata file consumed by Tesseract triggers this during legacy-classifier initialization, before any OCR output, gaining heap corruption, a crash (denial of service), or potentially controlled corruption. All deployments running Tesseract 5.5.3 or earlier are affected, particularly applications, document pipelines, and services that load traineddata files from untrusted or user-controlled sources. No fixed release is available as of this review, and no public proof-of-concept, KEV listing, or known in-the-wild exploitation exists.

Do: No patched version is available yet, so mitigate by ensuring Tesseract only loads .traineddata files from trusted sources and treat user-supplied traineddata as untrusted input. Deployments using only LSTM-based recognition can check whether their pipelines load legacy INTTEMP components, since the flaw is reached during legacy-classifier initialization. Monitor the tesseract-ocr GitHub repository for a fixed release and upgrade promptly once published.

CVE-2026-78155 is a privilege escalation flaw in the OnGres StackGres Kubernetes operator in which a low-privileged tenant who owns a database can elevate to administrator privileges. The CVSS vector (AV:N/AC:L/PR:L/UI:N/S:C) indicates a network-based attack requiring only low privileges and no user interaction, with impact crossing the security scope boundary between tenant and operator. A successful attacker gains high confidentiality, integrity, and availability impact at the operator level, effectively administrative control beyond their own database. Any organization running a StackGres-managed PostgreSQL cluster where tenants are granted database ownership is potentially affected. A public reference is tracked in the project's GitLab issue tracker (work item 3177); EPSS is 0.3% (20th percentile) and the flaw is not in CISA KEV, so no confirmed in-the-wild exploitation is currently known.

Do: Monitor the vendor's advisory (GitLab work item 3177) and upgrade the StackGres operator as soon as a patched release is published, since no fixed version is stated in the source data. Until then, limit database-ownership grants to trusted tenants and restrict network access to the operator's management interfaces, and review operator audit logs for unexpected privilege changes.

CVE-2026-76886 is a heap-based buffer overflow (CWE-122) in the C12.22 protocol dissector of the Wireshark network protocol analyzer, affecting versions 4.6.0 through 4.6.7 and 4.4.0 through 4.4.18. It is triggered when a vulnerable Wireshark or tshark instance dissects a maliciously crafted C12.22 packet, whether from live network traffic during a capture or from an attacker-supplied capture file. An attacker who can get such a packet into a capture session gains denial of service, crashing the analyzer and interrupting capture/analysis; the CVE is rated 9.8 (critical) under CVSS 3.1, although the described impact is limited to a dissector crash. Anyone running the listed Wireshark versions is affected, particularly analysts capturing on networks where third parties can inject traffic or processing untrusted capture files. Exploitation has not been reported in the wild; the bug is tracked in public GitLab issues, is not in CISA KEV, and EPSS estimates only a 0.3% probability of exploitation in the next 30 days.

Do: Upgrade Wireshark to the first release after 4.6.7 in the 4.6 branch or after 4.4.18 in the 4.4 branch as soon as the patch releases are available. Until then, disable the C12.22 dissector via Analyze > Enabled Protocols, and avoid running automated tshark/capture jobs or file-formatting analysis on untrusted traffic or capture files from untrusted sources.

jahlives openssl_encrypt versions before 1.4.0 contain a sandbox escape (CWE-95) in the IsolatedPluginExecutor component, which exposes Python type objects in the builtins available to code executed via restricted exec() calls. An attacker who can get crafted code into the sandbox (for example, by supplying a plugin) can traverse the Python class hierarchy via __class__.__mro__.__subclasses__() to reach and invoke system functions. This grants arbitrary operating-system command execution under the privileges of the host process; the flaw is rated 9.3 critical under CVSS 4.0 with a network attack vector and no privileges or user interaction required. Only deployments running openssl_encrypt older than 1.4.0 that execute untrusted code or plugins through IsolatedPluginExecutor are affected; version 1.4.0 and later are not. No exploitation in the wild is currently known (not in CISA KEV; EPSS is about 0.5%), but a public proof-of-concept reference exists in the vendor's GitHub security advisory GHSA-m25m-ggxg-239c.

Do: Upgrade openssl_encrypt to version 1.4.0 or later. Until upgraded, avoid running untrusted or third-party code through IsolatedPluginExecutor, or restrict the builtins exposed to exec() so Python type objects are not reachable. Review the vendor advisory (GHSA-m25m-ggxg-239c) and check whether your deployment executes user-supplied plugins inside the sandbox.

Wireshark versions 4.6.0 through 4.6.7 contain a heap-based buffer overflow (CWE-122) in the TTX Logger file parser, the component used to read TTX Logger capture files. The flaw is triggered when the parser processes a crafted or malformed TTX Logger file, and the CVSS vector (local attack vector, user interaction required) means an attacker needs a local user to open the malicious file. Successful exploitation crashes the application, producing a denial of service with no confidentiality or integrity impact. Anyone running Wireshark 4.6.0 through 4.6.7 is affected, especially users who open capture files obtained from untrusted sources. No in-the-wild exploitation is known: the issue is documented with a public PoC reference on the Wireshark GitLab tracker, EPSS is 0.1% (3rd percentile), and it is not in CISA KEV.

Do: Check the installed Wireshark version and upgrade to a release newer than 4.6.7 when available from wireshark.org. As an interim mitigation, do not open TTX Logger capture files from untrusted or unexpected sources with affected 4.6.x builds; the flaw only causes a crash, so risk is limited to availability.

WordPress Core contains an interpretation-conflict flaw (CWE-436) in the REST API batch endpoint that causes route confusion, allowing requests to be dispatched to an unintended route. When chained with the author__not_in WP_Query SQL injection (CVE-2026-60137), this allows a remote, unauthenticated attacker to inject SQL and escalate to remote code execution. Any WordPress site running 6.9.x before 6.9.5 or 7.0.x before 7.0.2 is affected, with no privileges or user interaction required (CVSS 3.1: 9.8). Exploitation is confirmed in the wild: the issue was added to CISA's KEV on 2026-07-21, EPSS is 97.3% (100th percentile), and a public PoC is reportedly fueling mass scanning under the 'wp2shell' label.

Do: Upgrade immediately to WordPress 6.9.5 (for 6.9.x) or 7.0.2 (for 7.0.x), or later, which fixes both this flaw and the chained SQL injection CVE-2026-60137; prioritize internet-facing sites because exploitation is unauthenticated and being mass-scanned. Review web and REST API logs for suspicious batch requests, SQL errors, unexpected admin users, or modified files. Organizations in scope for CISA BOD 26-04 must apply the required mitigations per vendor instructions or discontinue use.

PyMdown Extensions is a set of extensions for the Python-Markdown markdown project. From 10.0.1 until 10.21.3, pymdownx.snippets uses a string-prefix containment check in SnippetPreprocessor.get_snippet_path() in pymdownx/snippets.py when `restrict_base_path: True`, allowing markdown snippet directives to read files from sibling paths that share the same base_path prefix, such as docs and docs_internal. This is a regression of CVE-2023-32309. This issue is fixed in version 10.21.3.

Soup Sieve is a CSS selector library designed to be used with Beautiful Soup 4. Prior to 2.8.4, the CSS selector parser in soupsieve contains a regular expression vulnerable to catastrophic backtracking when processing an attribute selector with an unterminated quoted value in soupsieve/css_parser.py, allowing an attacker who can supply untrusted CSS selector strings to soupsieve.compile() or Beautiful Soup .select() / .select_one() to cause CPU exhaustion and denial of service. This issue is fixed in version 2.8.4.

Vitest is a testing framework powered by Vite. Prior to 3.2.5 and 4.1.0, the Vitest UI/API server on Windows used isFileServingAllowed incorrectly for /__vitest_attachment__, allowing \\?\\..\\ path traversal to read files outside the project; exposed API write and rerun features such as saveTestFile and rerun could also allow arbitrary script execution. This issue is fixed in versions 3.2.5 and 4.1.0.

decompress before 4.2.2 allows arbitrary symlink creation during archive extraction. When processing symlink entries (type === 'symlink'), the x.linkname field from the archive is passed directly to fs.symlink() without validation (index.js line 121). The preventWritingThroughSymlink check on line 98 only applies to file entries, not symlink creation. An attacker can craft an archive with symlink entries pointing to sensitive files outside the extraction directory (e.g., /etc/passwd), enabling information disclosure when the application reads the extracted contents.

UMTS FP protocol dissector crash in Wireshark 4.6.0 to 4.6.6 and 4.4.0 to 4.4.16 allows denial of service