Denial of service in FasterXML jackson-core's non-blocking (async) JSON parser allows unauthenticated network attackers to exhaust JVM heap and CPU by submitting JSON containing arbitrarily long number tokens that bypass the maxNumberLength constraint. Affected deployments are those using the async parser API - particularly reactive frameworks such as Spring WebFlux - running com.fasterxml.jackson.core:jackson-core 2.15.0-2.18.5 or 2.19.0-2.21.0, or tools.jackson.core:jackson-core 3.0.0-3.0.x. Publicly available exploit code exists (PoC JUnit test published in GHSA-72hv-8253-57qq); the vulnerability is not listed in the CISA KEV catalog at time of analysis.
SQL injection in the Database for Contact Form 7, WPforms, Elementor forms WordPress plugin (versions before 1.5.5) exposes the underlying WordPress database to unauthorized read access by any user holding a specific capability that defaults to administrators but can be explicitly delegated to lower-privileged roles. The CVSS scope-change rating (S:C/C:H) reflects that a successful injection reaches beyond the plugin itself into the shared WordPress database, potentially exposing all stored form submissions, user data, and other sensitive records. Publicly available exploit code exists per WPScan; however, EPSS at 0.16% (5th percentile) and absence from CISA KEV indicate no confirmed widespread active exploitation at time of analysis.
Stored cross-site scripting in the Brizy WordPress plugin before 2.8.19 allows any authenticated user with Contributor role or above to plant persistent XSS payloads via unsanitized featured-image focal-point coordinates submitted through an AJAX action. When a higher-privileged user (Editor or Administrator) later opens the affected post for review, the injected script executes silently in their browser session, enabling session hijacking or unauthorized administrative actions. A publicly available exploit exists per WPScan; the vendor has released a fix in version 2.8.19.
Stored Cross-Site Scripting in the PowerPress Podcasting plugin by Blubrry (WordPress) before version 11.16.11 allows any authenticated user holding the Contributor role to inject and persist arbitrary JavaScript via unsanitized Podcast Episode settings fields. The injected payload executes in the browsers of higher-privileged users - including editors and administrators - who view the affected content, enabling session hijacking, credential theft, or unauthorized administrative actions. A publicly available proof-of-concept exists per WPScan; no active exploitation is confirmed by CISA KEV, and the EPSS probability stands at a low 0.15%, suggesting opportunistic rather than widespread exploitation at this time.
Server-Side Request Forgery in the Visualizer WordPress plugin before 4.0.6 allows any authenticated Contributor-level user to pivot through the WordPress server and reach link-local instance-metadata endpoints (e.g., 169.254.169.254), returning the full response to the attacker. On cloud-hosted WordPress deployments (AWS, GCP, Azure), this enables non-blind exfiltration of IAM credentials and other instance metadata, potentially leading to full cloud account compromise. A publicly available proof-of-concept exploit exists; exploitation probability is currently low (EPSS 0.14%, 3rd percentile) but the concrete cloud-credential impact makes this a material priority for cloud-hosted sites.
Session fixation in Ghost Admin (v2.2.0-v6.54.0) allows an attacker who can plant a known session identifier in a victim's browser - via a secondary vulnerability on the same domain - to hijack an authenticated Ghost Admin session after the victim logs in. The root cause is that Ghost Admin's login flow did not call session regeneration, leaving the pre-login session ID valid post-authentication. Exploitation is chained and non-trivial, but the upstream fix (v6.54.1) is confirmed and straightforward to apply. No public exploit code and no CISA KEV listing identified at time of analysis.
Path traversal in Ghost CMS theme upload (CWE-22) allows an authenticated staff-level user to write arbitrary files outside the designated uploads directory, potentially altering the behavior of the entire Ghost installation. Affected versions span v0.10.0 through v6.54.0 of the npm/ghost package - a wide range covering essentially all modern deployments. No public exploit code or CISA KEV listing exists at time of analysis; however, the CVSS Scope:Changed metric reflects that a successful write can affect components beyond the theme storage boundary, elevating the real-world impact above the moderate base score suggests.
Catastrophic regex backtracking in Open WebUI 0.9.6-0.10.x allows any authenticated user to permanently stall a Uvicorn event-loop worker by supplying a crafted pattern to the built-in knowledge-search tool. The vulnerable `grep_knowledge_files` tool compiles attacker-supplied patterns with Python's unbounded `re` engine against attacker-uploaded file content, with no timeout anywhere on the execution path; a 30-character subject string measured at 74 seconds of CPU and a 40-character string extrapolates to roughly one day. No CISA KEV listing exists, but a complete proof-of-concept procedure is published in the GitHub security advisory GHSA-2f54-p244-32q6, making exploitation straightforward for any credential holder.
Tool source code disclosure in Open WebUI (≤ 0.10.2) allows any authenticated non-admin user who holds a read grant on a shared workspace tool to retrieve the full Python source of that tool via three unguarded API endpoints. The root cause is a Pydantic response-model inheritance defect where the `ToolUserResponse` subclass re-admitted the `content` field that the parent `ToolResponse` schema deliberately excluded, combined with a separate query-level defect in the list path. Because tool source commonly embeds hard-coded API keys, credentials, and internal service URLs, the practical confidentiality loss frequently exceeds the code itself; a detailed proof-of-concept is published in the GitHub security advisory, and the vendor-confirmed fix is available in 0.11.0 with no active exploitation (CISA KEV) recorded at time of analysis.
Unrestricted file upload in the Chat Widget WordPress plugin before 1.8.2 allows any network-reachable user to submit arbitrary files to the plugin's public response endpoint, bypassing all validation of type, extension, content, and size. Uploaded files are stored in the WordPress uploads directory under bare UUIDs - stripping the original extension - which explicitly prevents code execution or stored XSS despite the CWE-434 classification. Real-world impact is bounded to disk exhaustion and abuse of the hosting environment to serve attacker-controlled static content, conditional on the channel's response storage or mail-forwarding feature being configured. No public exploit identified via CISA KEV; however, a proof-of-concept is publicly available per WPScan, and EPSS at 0.16% (6th percentile) signals minimal observed exploitation activity.
Unauthenticated consent-record forgery and privacy-queue flooding affect GDPR Framework By Data443 WordPress plugin versions prior to 2.4.0. The plugin fails to verify the identity of the data subject or check authorization when recording cookie-consent choices and processing privacy requests, allowing any unauthenticated remote attacker to submit fraudulent GDPR consent records on behalf of arbitrary email addresses and to inject unlimited entries into the site's privacy-request queue. A publicly available proof-of-concept exists per WPScan; the vulnerability is not listed in the CISA KEV catalog, and EPSS sits at 0.15% (4th percentile), indicating low observed exploitation at time of analysis.
Server-side request forgery in Odysseus before commit 87babb5 allows admin-authenticated attackers to weaponize the embedding endpoint configuration API to probe internal network resources, including cloud instance metadata services at 169.254.169.254, RFC 1918 hosts, and other server-reachable targets. The vulnerability arises because the embedding endpoint accepted arbitrary URLs without validating scheme, host, IP range, or DNS rebinding protections, enabling partial response exfiltration from cloud credential endpoints and internal APIs. A publicly available exploit and blog post exist; the vulnerability is not currently listed in CISA KEV, but the high subsequent-system confidentiality impact (SC:H in the provided CVSS 4.0 vector) reflects real risk of cloud credential theft from metadata services.
Broken object-level authorization in the REST API Log WordPress plugin (all versions before 1.7.1) permits unauthenticated attackers who hold any valid log-download token - regardless of which entry it was issued for - to retrieve arbitrary log entries from any other user, exposing stored credentials, authentication tokens, and private REST API payloads. The root cause is twofold: tokens are not cryptographically bound to specific log entry IDs, and the plugin performs no WordPress capability check on the requester. A public exploit PoC is referenced via WPScan, though EPSS at 0.18% and SSVC exploitation status of 'none' indicate no confirmed widespread abuse at time of analysis.
Blind Server-Side Request Forgery in the EmbedPress WordPress plugin (all versions before 4.6.1) allows unauthenticated remote attackers to coerce the WordPress server into issuing arbitrary HTTP requests to internal hosts and services, bypassing WordPress core URL validation entirely. The vulnerability exists on unauthenticated endpoints, requiring zero credentials or user interaction, making mass exploitation trivially automatable against any site running a vulnerable plugin version. A publicly available proof-of-concept exists (via WPScan), though EPSS of 0.14% (4th percentile) suggests exploitation remains limited in practice; no active exploitation is confirmed by CISA KEV.
Unrestricted file upload in Trippo ResponsiveFilemanager (all versions through 9.14.0) exposes the dialog.php endpoint to unauthenticated remote attackers, enabling arbitrary file upload to the server. The product is end-of-life with no vendor support and no patch forthcoming; the vendor did not respond to disclosure. A public proof-of-concept exploit is available on GitHub, making exploitation accessible to a wide range of threat actors. The CVSS 4.0 vector rates the score at 6.9 (E:P), but the underlying CWE-434 class in a PHP application typically enables web shell deployment and full server compromise beyond what the assigned impact metrics suggest.
Path traversal in Ghost's database backup export endpoint allows Administrator-level users to remotely overwrite arbitrary files on the server filesystem, affecting all Ghost installations from version 1.20.1 through v6.54.0 (npm package pkg:npm/ghost). An authenticated admin can supply a crafted export filename containing path separators (e.g., '../../etc/crontab') when triggering a database export, causing Ghost to write the JSON export blob outside the intended directory to an attacker-controlled path. No public exploit code exists and this CVE does not appear in CISA KEV at the time of analysis; however, the CVSS Scope:Changed rating reflects that successful exploitation crosses the application's filesystem boundary, with integrity and availability consequences beyond Ghost's own data directory.
Improper authorization in KodBox 1.67 Build 02 allows unauthenticated remote attackers to bypass access controls via the msgWarning plugin endpoint, resulting in limited unauthorized access to confidential data. The CVSS 4.0 vector (PR:N/UI:N) confirms no authentication is required, and a public proof-of-concept exploit has been published on GitHub. The vendor did not respond to responsible disclosure, leaving no confirmed patch available at time of analysis.
Broken object-level authorization in Open WebUI's standard channel messaging API allows any authenticated channel member with write access to overwrite or permanently delete messages authored by other members. Affected versions are open-webui 0.5.0 through 0.10.2 (pip). The `update_message_by_id` and `delete_message_by_id` route handlers in `channels.py` enforced a write-permission check for standard channels instead of an authorship check, meaning the same credential that grants posting rights silently granted full edit and delete authority over every other member's messages in that channel. A proof-of-concept script demonstrating end-to-end exploitation was included in the report; no public exploit code is independently confirmed at time of analysis, and the vulnerability is not listed in the CISA KEV catalog.
Horizontal privilege escalation in the Paid Membership Subscriptions WordPress plugin (all versions before 3.0.8) allows any subscriber-level authenticated user to hijack and overwrite another member's subscription - including plan tier, status, and expiration date - by submitting a crafted change-subscription checkout request with an arbitrary subscription ID. The root cause is a missing ownership verification check at the business logic layer of the checkout flow (CWE-284). A publicly available proof-of-concept exploit has been published by WPScan, though EPSS sits at just 0.13% (3rd percentile), suggesting limited automated exploitation so far. The vulnerability is not listed in CISA KEV at time of analysis.
Server-Side Request Forgery in Simple Google Calendar Outlook Events Widget WordPress plugin (all versions before 3.1.0) permits unauthenticated remote attackers to supply arbitrary URLs that the server fetches without validation, potentially exposing responses from internal services. The plugin is confirmed vulnerable by WPScan and a public exploit exists, though EPSS at 0.14% (4th percentile) indicates limited observed exploitation despite the available POC. This vulnerability is not listed in CISA KEV at time of analysis.
Cross-site scripting in Perspective 5.0.0's built-in Debug plugin enables an authenticated attacker with low privileges to inject arbitrary HTML and JavaScript into the embedding page's origin by writing crafted table cell values that bypass RFC 4180 quoting. The Debug plugin's CSV serialization renderer assigns table cell content directly to innerHTML without sanitization, meaning event handler attributes and unquoted injection payloads execute as script in any page embedding the component. A publicly available proof-of-concept exists; the vulnerability is not listed in CISA KEV, but the low attack complexity and available exploit code elevate practical risk above the moderate CVSS 4.0 score of 5.1.
Cross-site scripting in Ghost CMS (versions 5.26.0 through 6.54.0) allows a high-privileged administrator to plant arbitrary JavaScript via the Universal Import feature, which fails to sanitize the `feature_image_caption` field and link card URLs before storing and rendering them. The fix in v6.54.1 introduces DOMPurify sanitization across both the Ember-based admin post history modal and the Node.js frontend rendering pipeline, alongside a strict URL protocol allowlist in the JSON-to-HTML importer. No active exploitation is confirmed (not in CISA KEV) and no public proof-of-concept is known; the high-privilege prerequisite substantially limits the realistic attacker pool.
Blind password hash disclosure in Ghost Admin API versions prior to 6.54.1 allows any authenticated staff-level user to retrieve hashed passwords of other staff accounts by crafting filtered API queries against the posts or pages browse endpoints. The root cause is insufficient field-level access control in MongoDB filter queries, exposing internal authentication fields (CWE-200). No active exploitation is confirmed (not in CISA KEV), and downstream account takeover is partially mitigated by Ghost's Device Verification and email MFA features, though the hash disclosure itself is confirmed and unambiguously fixed in 6.54.1.
Ghost CMS subscription checkout accepts redemption of archived offers due to missing offer-status validation in the member router controller, affecting all self-hosted Ghost instances from v4.22.0 through v6.54.0. Unauthenticated users who possess or can discover an archived offer ID can submit a valid checkout request and obtain unauthorized subscription discounts or access at a revoked price point. No public exploit code has been identified and this CVE is absent from CISA KEV, but the direct financial impact on subscription-based Ghost sites makes it a meaningful business-logic risk for affected operators.
Stored cross-site scripting in the Quiz and Survey Master (QSM) WordPress plugin before 11.2.2 allows contributor-level (and higher) WordPress users to inject arbitrary JavaScript into quiz question settings via an unquoted HTML attribute. The injected payload persists and executes in the browser of any site visitor who views the affected quiz, enabling session hijacking, credential theft, or defacement targeting the broader user base. No active exploitation is confirmed by CISA KEV, but a public proof-of-concept is available via WPScan, and vendor-released patch version 11.2.2 is available.
Stored cross-site scripting in the Nested Pages WordPress plugin (all versions before 3.2.15) allows authenticated users holding the Editor role - or Contributor/Author roles when the plugin is configured to manage those post types - to inject arbitrary JavaScript via unsanitized post titles that are reflected unescaped into HTML attributes on the plugin's administrative listing screen. Any higher-privileged user (including Administrators) who subsequently loads that screen will execute the attacker-controlled script within their own authenticated session, enabling session hijacking, credential theft, or unauthorized administrative actions. A publicly available proof-of-concept exploit exists per the WPScan advisory; the vulnerability is not listed in CISA KEV and the EPSS score is negligible (0.15%, 5th percentile), suggesting opportunistic rather than widespread active exploitation.
Open redirect in Clearfy Cache WordPress plugin (versions before 2.4.3) allows unauthenticated attackers to redirect site visitors to arbitrary external URLs by exploiting the Cyrlitera old-URL redirect handler, which passes a decoded request URI to an unsafe redirect function without target validation. Exploitation requires that the non-default Cyrlitera redirect option is enabled on the target site, and a victim must interact with an attacker-crafted URL. No public KEV listing exists, EPSS sits at the 4th percentile (0.14%), but a publicly available proof-of-concept has been published by WPScan.
Improper access control in the wpForo Forum WordPress plugin before 3.1.3 permits any subscriber-level authenticated user to write to administrator-controlled profile fields - including account activation status and forum reputation score - during a standard self-service profile update. This allows a banned or pending user to self-activate their own account, bypassing moderator enforcement, and to forge an arbitrarily high reputation score that may unlock elevated forum privileges. No CISA KEV listing exists, but a publicly available exploit is documented by WPScan; the EPSS score of 0.14% (4th percentile) indicates low observed exploitation probability at scale, though the low barrier to attack (any registered user) makes it a meaningful integrity risk for moderated forums.
Missing authorization in the Wired Impact Volunteer Management WordPress plugin (versions before 2.8.2) allows any authenticated WordPress user with at least a Subscriber role to remove RSVPs belonging to other users from any volunteer opportunity. The vulnerability stems from an AJAX action handler that neither verifies the requesting user's authorization level nor confirms ownership of the RSVP being deleted, enabling unauthorized data manipulation across the entire volunteer management system. A publicly available proof-of-concept exists via WPScan, though EPSS sits at 0.13% (3rd percentile) and SSVC confirms no active exploitation, suggesting limited real-world attacker interest at this time.
Clearfy Cache WordPress plugin versions before 2.4.3 exposes admin-only settings pages and their contents - including administrative nonces - to any authenticated user at Subscriber level or above due to a missing capability check on an alternate admin-page dispatch path. While the canonical admin URL enforces access controls correctly, the unguarded dispatch path bypasses those restrictions entirely. A publicly available proof-of-concept exists per WPScan reporting, though EPSS sits at 0.13% (3rd percentile), indicating low observed exploitation probability despite POC availability; no confirmed active exploitation (CISA KEV) has been recorded.
Contest Gallery WordPress plugin (all versions before 30.0.7) exposes the site's complete stored OpenAI prompt history to any authenticated user holding at minimum a Subscriber role, due to a handler that performs neither a capability check nor a nonce verification. An attacker with a basic WordPress account can call the unprotected AJAX endpoint directly to exfiltrate the full prompt history, potentially exposing confidential business logic, AI workflows, or sensitive data embedded in prompts. No active exploitation is confirmed (not in CISA KEV), but a publicly available proof-of-concept exists per WPScan, and EPSS remains low at 0.13% (3rd percentile).
Missing authorization in the miniOrange 2FA WordPress plugin before 6.2.7 allows any authenticated low-privileged user to trigger OTP email delivery to arbitrary email addresses and exhaust the site's metered OTP quota. The plugin's second-factor enrollment endpoint neither restricts which users may invoke it nor binds the OTP recipient to the enrolling user's own address, enabling both unsolicited OTP email delivery and a denial-of-service attack against other users' 2FA flows. No public exploitation or KEV listing has been identified, though a public proof-of-concept is available via WPScan; EPSS stands at 0.13% (3rd percentile), indicating low observed exploitation activity.
Client-side SSRF in Open WebUI (pip package versions 0.6.34 through 0.10.x) allows any authenticated user who can post chat content to redirect victims' browsers into issuing arbitrary outbound GET requests against internal hosts, same-origin endpoints, or CORS-permissive targets. The vulnerability stems from the Vega visualization renderer accepting its default resource loader against untrusted chat-embedded chart specs, providing two independent fetch paths - `data.url` at view construction and image-mark `url` attributes emitted into the rendered SVG - both entirely invisible to server-side SSRF controls. No public exploit was identified in CISA KEV, but a working proof-of-concept was published in the GitHub security advisory itself (GHSA-rffm-9q57-q649), significantly lowering the exploitation bar.
Denial-of-service via recursive GEOMETRYCOLLECTION parsing crashes Django processes through a segmentation fault in the underlying GEOS library. Django's GEOSGeometry class, GeometryField form field, and spatial field lookups in versions 5.2, 6.0, 6.1 (RC), and main are all affected when processing WKT or WKB input containing deeply nested or excessively numerous GEOMETRYCOLLECTION objects. No public exploit code or active exploitation has been identified at time of analysis; the Django team rated this 'moderate' severity and issued fixed releases Django 5.2.17 and 6.0.8 on August 4, 2026.
Memory exhaustion in Django's `check_for_language()` function allows denial-of-service via the `set_language` view when it is explicitly enabled in URL configuration. Attackers can submit many distinct, very long language codes via POST requests, causing each code to be cached as an in-memory key and progressively consuming process memory. Django rates this severity 'low' because built-in guards - `DATA_UPLOAD_MAX_MEMORY_SIZE` and the cache's maximum entry count - bound the total exploitable memory; no public exploit identified at time of analysis.
Unauthenticated access to the network setup wizard API in H3C NX15 firmware V100R017 allows remote attackers to interact with the /api/wizard/networkSetup endpoint without presenting valid credentials. The root cause is CWE-306 - missing authentication for a critical function - meaning the device exposes a privileged configuration path with no access control gate. A public proof-of-concept repository exists on GitHub, confirming exploitability; no CISA KEV listing was identified at time of analysis, suggesting exploitation has not yet been confirmed at scale.
Unauthenticated remote clients can permanently exhaust the global monitored-item quota in Eclipse Milo OPC UA SDK versions 1.0.0 through 1.1.4, effectively preventing all clients from creating new monitored items until the server is restarted. The attack combines two defects: the server's ExtensionObject decoder lacks recursion-depth enforcement, enabling deeply nested PubSub structures in a CreateMonitoredItems event filter to trigger a JVM StackOverflowError, and the quota accounting code is not exception-safe against unchecked JVM Errors, so the reserved slot is never reclaimed. No public exploit code has been identified and this CVE is absent from the CISA KEV catalog; however, the CVSS 4.0 vector AV:N/AC:L/PR:N/UI:N confirms the attack is network-reachable with no authentication or complexity barriers.
Session security diagnostics nodes in Eclipse Milo 0.6.0 through 1.1.4 expose sensitive OPC UA session metadata without enforcing any authorization checks, a CWE-862 (Missing Authorization) flaw. An unauthenticated anonymous client connecting over a None/None (no-security) endpoint can enable the diagnostics subsystem with no certificate or credentials whatsoever; separately, a client presenting a trusted application certificate over SignAndEncrypt can read the SessionSecurityDiagnosticsArray for all active server sessions, harvesting usernames, login history, authentication mechanisms, session security modes and policies, and public client certificates. No public exploit code exists and the vulnerability is not listed in CISA KEV, but the zero-authentication path and prevalence of None/None endpoints in legacy OT/IIoT environments meaningfully lower the exploitation barrier.
Memory corruption in Qualcomm Snapdragon's registry value processing exposes devices to potential code execution by high-privileged local attackers. The flaw, rooted in CWE-120 (classic buffer overflow), is triggered when a direct query method processes registry values supplied with incorrect types - bypassing proper type-length validation and corrupting adjacent memory. While the full C:H/I:H/A:H impact triad indicates serious consequences if reached, the PR:H (high privilege) and AV:L (local) CVSS metrics substantially narrow the realistic attack surface. No public exploit code exists and this CVE is not listed in CISA KEV at time of analysis.
Captive portal authentication bypass in Zyxel WAX650S firmware exposes guest or restricted wireless networks to unauthorized access by any WLAN-adjacent attacker. The flaw resides in the social_login.cgi CGI handler, where CWE-287 (Improper Authentication) allows requests to circumvent captive portal enforcement entirely - not just weaken it. All firmware versions through 7.10(ABRM.4)C0 are confirmed affected per the vendor's own advisory. No public exploit code or CISA KEV listing has been identified at time of analysis, though the low attack complexity makes opportunistic exploitation plausible once the mechanism is understood.
Path traversal in HUMANIST Digital Human Resources version 26.0 enables authenticated remote attackers to read arbitrary files from the server filesystem without any user interaction. The CVSS vector (AV:N/AC:L/PR:L/UI:N) confirms network-accessible exploitation requiring only a low-privilege account - consistent with a standard HR system employee login - with high confidentiality impact and no integrity or availability loss. No public exploit code and no active exploitation (CISA KEV) have been identified at time of analysis; the vulnerability was disclosed by TR-CERT (Turkish national CERT) under advisory TR-26-0737.
Stack-based out-of-bounds read in stunnel's s_vlog logging function allows remote unauthenticated attackers to crash the service by sending protocol inputs that generate log messages exceeding 1024 bytes via vsnprintf. The primary impact is service disruption (availability), with a limited integrity impact in corner cases where trailing newline characters in the stack buffer can be overwritten with null bytes. No public exploit code has been identified at time of analysis, and this CVE is not listed in the CISA KEV catalog.
Open WebUI versions 0.9.0 through 0.10.x expose an instance-wide availability collapse through unbounded iCalendar recurrence rule parsing in the automation scheduler. The `_parse_rule` function in `backend/open_webui/utils/automations.py` anchors MINUTELY and HOURLY rules to a hardcoded 2000-01-01 epoch and iterates forward one interval at a time on the synchronous ASGI event loop - a single RRULE:FREQ=MINUTELY rule consumes approximately 18.9 seconds of blocking CPU, denying service to every concurrent user. Critically, the stall is not transient: once the automation is stored, the scheduler recomputes the next-run value on every 10-second poll, permanently wedging the instance until it is restarted or the automation is surgically removed. No public exploit identified at time of analysis, though the vendor advisory includes precise proof-of-concept timing measurements that make the attack trivially reproducible.
Insufficient session expiration in HUMANIST Digital Human Resources version 26.0 allows an attacker who obtains a valid session token to replay it indefinitely, effectively hijacking an authenticated user's session and gaining unauthorized access to HR data. Reported by TR-CERT (TR-26-0737), the flaw affects all 26.0 builds prior to 26.1 and carries a high confidentiality impact due to the sensitive nature of human resources data processed by the platform. No public exploit code has been identified and this vulnerability is not listed in the CISA KEV catalog.
Race condition during LoRA manager singleton initialization in NVIDIA Dynamo for Linux (all versions through v1.1.0) allows unauthenticated remote attackers to cause denial of service or corrupt inference data by exploiting a TOCTOU flaw in concurrent initialization paths. The CVSS vector assigns high availability impact alongside low integrity impact, reflecting that service crashes are the primary outcome while data tampering represents a secondary, harder-to-achieve consequence. No public exploit code has been identified at time of analysis, and CISA has not added this to the Known Exploited Vulnerabilities catalog.
Race condition in NVIDIA Dynamo's LoRA manager singleton initialization exposes Linux-based AI inference deployments running versions up to v1.2.0 to data tampering and denial of service. The flaw, classified under CWE-362, arises during concurrent initialization of the singleton responsible for managing Low-Rank Adaptation (LoRA) adapters, allowing competing threads or network-triggered requests to corrupt shared state before synchronization is established. No public exploit code or CISA KEV listing has been identified at time of analysis; however, the network-accessible attack vector paired with no privilege requirement makes this relevant for any internet-facing or multi-tenant Dynamo deployment.
Information disclosure in Firefox for Android and Firefox Focus for Android exposes sensitive user data to network-based attackers who can lure a victim into interacting with a crafted web resource. The vulnerability, tracked under CWE-200 and assigned a CVSS score of 6.5 with high confidentiality impact, is limited to Android platform builds and does not affect desktop Firefox. Mozilla patched this in Firefox 153.0.3 per advisory mfsa2026-73; no public exploit or active exploitation has been identified at time of analysis.
Use-after-free in Zephyr RTOS 4.4.0's Bluetooth ATT host layer (subsys/bluetooth/host/att.c) allows an adjacent unauthenticated peer to crash the Bluetooth host or potentially corrupt live memory by disconnecting while an ATT PDU is in-flight. The tx_meta_data_storage[] array retains a dangling att_chan pointer after L2CAP tears down and frees the channel slab slot, and the deferred att_tx_destroy_work_handler running on the system workqueue dereferences that freed pointer. No pairing or special privileges are required - a routine ATT connection is sufficient to reach the vulnerable code path. No public exploit has been identified at time of analysis, and this vulnerability is not listed in the CISA KEV catalog.
Session flow control enforcement in Apache Qpid ProtonJ2 (versions through 1.1.0) can be bypassed by an authenticated attacker who sends data exceeding the negotiated incoming flow control window, potentially causing denial of service on the receiving side. The AMQP session-level window is a back-pressure mechanism; failure to enforce it can lead to unbounded memory consumption or processing overload in the message broker or client. No public exploit code and no CISA KEV listing have been identified at time of analysis. The vendor-released fix is version 1.2.0.
Unbounded echo flow responses in Apache Qpid Broker-J (all versions through 10.0.1) allow an authenticated AMQP 1.0 client to exhaust broker resources and cause denial of service by flooding the broker with unanswered echo flow frames. The root cause is absent rate-limiting on echo flow response logic in the qpid-broker-plugins-amqp-1-0-protocol plugin. No public exploit code or CISA KEV listing has been identified; the vendor has released a fix in version 10.1.0.