Fabio 1.6.6 through 1.7.1 (fixed in 1.7.2) fails to protect operator-renamed trust headers, so an unauthenticated remote client can force Go's reverse proxy to strip the configured ClientIPHeader, TLSHeader, or RequestID from its own request by naming them in the inbound Connection header, causing backends that rely on those signal headers for authorization or audit logging to lose them. The exposure is narrow: the three affected options are empty by default, the hardcoded forwarded-header allowlist (X-Forwarded-*, Forwarded, X-Real-Ip, X-Request-Id, X-Client-Ip) remains protected, and impact only materializes in deployments that both enable these options and have backends that trust the injected values. No public exploit code was identified at time of analysis and the issue is not in CISA KEV, though the integrity weakness is remotely reachable with no authentication (CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:N/I:L/A:N, 5.8).
Reflected cross-site scripting (XSS) in CuteNews 2.1.2 allows the value of the HTTP Referer header to be copied, unescaped, into the HTML response generated when POST messages are submitted to index.php, so a crafted Referer value can execute JavaScript in a victim's browser under the site's origin. No authentication is required (PR:N), but exploitation depends on user interaction (UI:R) and on an attacker-controlled or same-site intermediary page that shapes the Referer value, which is why the assessed vector carries AC:H; browsers set Referer automatically rather than from attacker-supplied links, so a passive victim cannot be triggered without the crafted delivery context. Independent assessment rates this a genuinely low-priority reflected XSS, EPSS is only 0.14% (4th percentile), and no public exploit code has been identified at time of analysis; there is no confirmed CISA KEV active exploitation.
Host-key trust confusion in Warpgate's browser SSH client (warpgate-web-ssh) lets a malicious or compromised SSH jump host have its host key stored against the final target's address, so a machine later presenting that key at the target address is accepted as the intended destination - permitting interception of user traffic and, when certificate authentication is used, issuance of a fresh certificate to the impersonating host. Affected are Warpgate 0.25.0 through 0.27.5 on the Web SSH path only; the native SSH proxy tracks each hop separately and is unaffected. Exploitation requires a high-barrier chain (Web SSH session by an authenticated user per PR:H, a target reached through a jump/proxy hop, Prompt or AutoAccept host-key mode with UI:R in Prompt mode, attacker possession of the jump host's private key, and a man-in-the-middle position at the target), and no public exploit code has been identified at time of analysis, with the issue not appearing in CISA KEV.
A double-free (CWE-415) in Zephyr's experimental USB host stack lets a malicious or malformed USB peripheral crash the device: the stack frees the per-device configuration-descriptor buffer udev->cfg_desc on three error paths in usbh_device_set_configuration() but never clears the pointer, so the guarded cleanup in usbh_device_free() frees the same chunk again. Only builds with CONFIG_USB_HOST_STACK enabled are exposed, and because that stack is experimental and off by default most Zephyr deployments are unaffected; exploitation requires physical access to plug in the attacker-controlled device and, on v4.1.0-v4.3.x, a device-initiated removal or duplicate-connect event after a failed enumeration. With the default CONFIG_SYS_HEAP_HARDENING_BASIC the double free is caught and triggers k_panic() - a deterministic denial of service - while builds using CONFIG_SYS_HEAP_HARDENING_NONE or older releases corrupt the free list of the small 1024-byte usb_device_heap instead. No public exploit identified at time of analysis and no CISA KEV listing; the vendor has published a fix (commit d46a8c1f) and a GitHub security advisory, making this a genuine but low-priority issue given the physical-access and build-configuration prerequisites.
Information disclosure in ScadaBR up to and including version 1.1 allows an attacker to pull the complete project export - data sources, data points and graphical views - by invoking the unprotected EmportDwr.createExportJSON DWR method through /ScadaBR/export_project.htm. The vendor's CVSS 4.0 vector lists PR:N, but the fix adds Common.getUser() plus Permissions.ensureAdmin(), which indicates the endpoint executes inside an authenticated session, so realistic exploitation requires a valid low-privilege ScadaBR account rather than true anonymous access; impact is confined to read-only configuration disclosure with no integrity or availability effect. Publicly available exploit code exists and the issue is fixed in ScadaBR 1.2.0, though there is no confirmation of active exploitation.
SQL injection in Drogon (drogonframework drogon) up to version 1.9.13 allows remote attackers to manipulate the ORM filter argument in makeCriteria (orm_lib/src/Criteria.cc), potentially reading or altering database contents. Exploitation requires a Drogon-based application that forwards attacker-controllable input into the ORM Criteria filter without parameterization; applications that do not expose this argument to user input are not vulnerable. Publicly available exploit code exists, no vendor patch has been released, and the vendor did not respond to disclosure.
Keycloak's Admin REST API allows authenticated administrators with delegated privileges over a realm to read or modify sensitive client configurations in the master realm by abusing a per-request in-memory cache that resolves clients by unique identifier without verifying realm ownership. The flaw affects Red Hat Build of Keycloak and Red Hat Single Sign-On 7, requires the attacker to hold privileged delegated administration rights and to know or reference a target client's unique ID, and is assessed as a moderate-priority authorization issue rather than a high-CVSS critical defect. No public exploit code was identified at time of analysis, and successful exploitation could expose client credentials or redirect administrative login attempts to malicious sites.
Local information disclosure in the Matter Chip (connectedhomeip) controller lets an attacker who already holds low privileges on a commissioning host read sensitive data through the PerformCommissioningStep code path in ChipDeviceController.cpp. The issue is rated CVSS 5.5 (AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:N) and is confidentiality-only, so the attacker cannot modify or disrupt the device via this flaw. EPSS is low (0.15%, 4th percentile) and there is no KEV listing or confirmed public exploit code; overall signals point to a genuine but low-urgency issue.
A stack buffer overflow in libstoragemgmt's SCSI VPD page 0x80 serial-number parser lets an attacker who controls a local or virtual storage device crash or destabilize the host process that enumerates that device, yielding a denial of service with no confidentiality or integrity impact (CVSS 3.1: AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H, base 5.5). Exploitation requires low privileges on the host (PR:L, so an authenticated/local actor) plus the ability to attach or present a device, and it only triggers when the affected process actually queries the device's serial number - the vector is local, not remotely reachable over a network. No public exploit code was identified at time of analysis and there is no confirmed active exploitation (CISA KEV); the practical risk is concentrated in environments that enumerate untrusted or attacker-attachable storage, such as multi-tenant virtualization or hosts that accept hostile removable, iSCSI, or hypervisor-presented disks.
Local information disclosure in Dell Command | Integration Suite for System Center (versions prior to 6.7.2) allows a low-privileged, authenticated local user to read sensitive data that should not be accessible to that account. The root cause is incorrect default file or directory permissions (CWE-276) shipped with the SCCM add-on, which grant broader read access than intended. Exploitation requires an existing local session on the Windows host running the affected software - there is no remote vector, no user interaction, and no privilege escalation needed beyond the initial foothold - and the result is confidentiality impact only, with no code execution, data modification, or availability impact. There is no public exploit identified at time of analysis and the flaw is not listed in CISA KEV; the assessed CVSS 3.1 base score is 5.5 (AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:N).
Local information disclosure in Dell Command | Intel vPro Out of Band versions prior to 4.7.2 lets an authenticated, low-privileged user on a Windows host read data that should have been restricted, because the product ships files, directories, registry keys, or objects with overly permissive default access controls (CWE-276). Per the vendor's own advisory the impact is limited to confidentiality - no integrity or availability effect - and the vector requires local access plus a valid low-privilege account (AV:L/PR:L/UI:N), which is why it scores a moderate 5.5. No KEV listing or public exploit code was identified at time of analysis, and no public exploit has been identified; real-world risk is confined to hosts where an attacker already has a foothold, so this is routine patch-cycle work rather than an emergency.
Open redirect in Tautulli prior to 2.17.2 lets an unauthenticated attacker redirect a victim off-site via the /auth/redirect endpoint, enabling phishing and post-login redirect-flow abuse; the CVSS vector (AV:N/AC:L/PR:N/UI:R) confirms no authentication is needed but the victim must click a crafted link. Only instances running the default root HTTP_ROOT configuration are exploitable - deployments with a custom non-root HTTP_ROOT path prefix are not affected. No public exploit code has been identified at time of analysis and the issue is not listed in CISA KEV, so this is a moderate-severity, low-priority targeted-phishing risk rather than a mass-exploitation threat; a vendor-released patch is available in 2.17.2.
Authenticated workspace members holding only the lowest MaxKB role can bypass the per-tool denial enforced by WorkspaceUserResourcePermission and execute tools they were explicitly denied in versions 2.0.0 through 2.9.2. The bypass works by binding a denied tool's identifier through tool_ids, skill_tool_ids, or mcp_tool_ids on the agent or workflow dispatch path, which does not re-apply the grant check that the dedicated tool routes perform, and because execution decrypts server-side init_params the caller can receive credentials carried by the denied tool. The flaw requires a valid low-privilege account (CVSS vector PR:L, unauthenticated attackers are not sufficient) and knowledge or enumeration of the denied tool's identifier; there is no public exploit identified at time of analysis and no vendor patch, with the published score of 5.4 understating confidentiality impact relative to our assessment of CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:L/A:N.
MaxKB 2.10.2-lts and earlier accepts expired non-permanent ApplicationApiKey credentials on the /chat/api/mcp endpoint, allowing an authenticated attacker who already holds such a key secret to initialize the MCP interface, enumerate the published application's tool metadata, invoke tools/call, and persist application_chat and application_chat_record records - operations the standard /chat/api/open path refuses with the same key. The root cause is CWE-613 (Insufficient Session Expiration): the MCP authentication path validates only the key secret and its active flag while omitting the is_permanent and expire_time checks enforced elsewhere. This is a bounded credential-lifecycle bypass rather than unauthenticated access - the attacker must already possess a leaked, retained, or previously delegated key that is still marked active and past its expiry - and no public exploit has been identified at time of analysis, with no vendor-released patch identified at time of analysis.
Missing authorization checks in the MaxKB enterprise AI assistant (2.10.2-lts and earlier) let an authenticated normal user in the same workspace harvest another user's application_id from the homepage application question-ranking endpoint, then place that identifier into an application-node of an attacker-owned published workflow application to obtain output generated by the victim application and create durable application_chat and application_chat_record rows under it. Exploitation is constrained: the victim application must be published and show ranking activity in the attacker-selected date range, the attacker must know or guess its name, and direct application-detail and debug-open routes continue to deny access - only the ranking endpoint leaks the identifier - and it requires a valid workspace account (CVSS 3.1 AV:N/PR:L, score 5.4). No public exploit identified at time of analysis and no fixed version is available as of this review.
Broken authorization in HomeBox prior to 0.26.0 lets any authenticated low-privilege user force an arbitrary account into their own group by POSTing a victim's userId to /v1/groups/members, with no owner role, invitation token, victim consent, or victim notification required. The direct impact is unauthorized group enrollment plus disclosure of the target's email address and name through the resulting member list (CVSS 5.4, CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:L/A:N), and the forced membership also satisfies the prerequisite for chaining a separate cross-group inventory-wipe issue. No public exploit code was identified at time of analysis and the flaw is not in CISA KEV; the endpoint and handler were removed entirely in the 0.26.0 fix, so patched instances are not exploitable.
Broken object-level authorization (IDOR) in MaxKB 2.0.0 through 2.10.2-lts lets an authenticated workspace user read or modify documents and paragraphs that belong to another user's knowledge base by sending an attacker-owned knowledge_id in the request path together with a victim's document_id or paragraph_id. The affected operate routes authorize only the knowledge_id and never verify that the referenced object actually belongs to that knowledge base, so a low-privileged account plus a known victim object UUID is sufficient, with no user interaction required (CVSS 3.1: AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:L/A:N; base score 5.4). No public exploit code has been identified at time of analysis and no vendor-released fixed version is available, so real-world risk is moderate and firmly gated by the requirement for an authenticated account on the same deployment.
Stored cross-site scripting in Discourse before 2026.7.0 allows an authenticated user who can set a custom display name to persist markup that is later rendered in post action descriptions, so that any other user browsing the affected activity streams runs attacker-controlled script in their browser. The vendor scores it 5.4 (Medium) with the vector CVSS:3.1/AV:N/AC:L/PR:L/UI:R/S:C/C:L/I:L/A:N, reflecting an authenticated attacker (PR:L) and a required victim interaction (UI:R), with scope change because the script executes in other users' sessions against the Discourse origin. The vendor-released fix is version 2026.7.0; there is no public exploit code identified at time of analysis and no CISA KEV listing.
Stack-based buffer overflow in fetchmail when built with NTLM support allows a malicious or compromised mail server to trigger memory corruption by sending a crafted NTLM Type 2 challenge, potentially leading to remote code execution, authentication failure, or process termination. Exploitation requires fetchmail to be built with NTLM and configured to authenticate via NTLM against an attacker-controlled or man-in-the-middle'd server; no public exploit code or active exploitation has been identified. The flaw carries a CVSS score of 8.1 with high confidentiality, integrity, and availability impact, though reliable RCE is limited by stack-frame layout and modern memory hardening (stack canaries, ASLR, FORTIFY_SOURCE) on RHEL builds.
Process termination in Envoy Proxy affects versions prior to 1.36.10, 1.37.6, 1.38.4, and 1.39.1 when a scoped IPv6 address (for example fe80::1%1) reaches either the ORIGINAL_DST transparent-proxy cluster path or a QUIC client connection undergoing port migration; the address is re-serialized with its percent scope suffix and then handed to inet_pton, which cannot parse it, so an EnvoyException is thrown and the process aborts. The assessed vector is CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:H (5.3), meaning an unauthenticated remote condition can be reached without user interaction but only under narrow, topology-dependent prerequisites, and impact is availability-only - there is no confidentiality or integrity loss. This is a genuine but conditional availability-only issue: it is not a default configuration (transparent proxying with ORIGINAL_DST, or QUIC, must be in use and the destination must actually be a scoped/link-local IPv6 address), no public exploit identified at time of analysis, no CISA KEV listing, and no EPSS signal was supplied with the data.
Path-confusion access-control bypass in Envoy Proxy before versions 1.36.10, 1.37.6, 1.38.4, and 1.39.1 allows remote, unauthenticated clients to reach resources that a path-based Envoy policy (path-based route selection or path-based authorization) intended to protect. Envoy matches the raw request path, while servlet backends such as Apache Tomcat strip semicolon matrix parameters from each path segment before resolving the resource, so a request like /admin;x=1 is matched by Envoy against an unprotected fallback route while the backend still serves the protected resource; Envoy's legacy ignore_path_parameters_in_path_matching option only truncates at the first semicolon and does not reproduce per-segment backend behavior. Exploitation is conditional (it requires both a path-based Envoy decision and a backend that strips semicolon parameters per segment) and the assessed impact is limited to Low confidentiality with no integrity or availability effect; no public exploit identified at time of analysis.
Path confusion in Envoy Proxy lets remote, unauthenticated clients bypass path-based authorization and routing policy by attaching semicolon parameters to dot and dotdot segments - a request such as /user/..;foo=bar/admin is not canonicalized to /admin even when path normalization is enabled, so a client can present one path to Envoy's policy engine and a different effective path to the upstream. The security consequence materializes only when Envoy itself makes a path-based RBAC or routing decision on the uncollapsed value, or when a downstream/upstream collapses the segment per RFC 3986 while Envoy does not; without such a mismatch the exposure is limited to low-confidentiality impact. The issue is fixed in Envoy 1.36.10, 1.37.6, 1.38.4, and 1.39.1; no public exploit code was identified at time of analysis and no CISA KEV or EPSS signal was provided. Note that the attached tag reads 'Authentication Bypass', but the assessed impact is a moderate-severity access-control (path confusion) bypass, not an authentication bypass.
Unauthenticated information disclosure in InvenTree prior to 1.4.0 lets a remote caller retrieve plugin names, setting keys, descriptions, types, choices and non-secret configuration values through /api/plugin/settings/ and the per-plugin settings endpoints, because PluginSettingList, PluginAllSettingList and PluginSettingDetail registered GlobalSettingsPermissions without the IsAuthenticated permission used by equivalent global-settings endpoints. Exposure is confined to metadata and non-secret configuration: protected secret values remain masked as three asterisks, only safe methods are readable, and at least one plugin must be installed and active for the disclosure to be meaningful, so the practical impact is limited. The assessed CVSS 3.1 vector is AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N (score 5.3), confirming remote unauthenticated exploitation with low confidentiality impact and no integrity or availability effect; no public exploit identified at time of analysis.
Authorization bypass in GoCD versions 18.7.0 through before 26.1.0 lets an authenticated user with a lower-privileged account read a pipeline template's configuration without the required view permission. The Get Template Config API selects its authorization filter through a case-sensitive comparison of the HTTP method name, so a request sent with nonstandard capitalization (for example 'Get' instead of 'GET') slips past the permission check; per the assessed CVSS 3.1 vector (AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:N/A:N) exploitation requires only a valid low-privilege account and exposes read-only configuration data, with no modification of data and no exposure of encrypted secure variables. No public exploit identified at time of analysis; the vendor-released fix is GoCD 26.1.0, and deployments whose reverse proxy rejects or normalizes non-uppercase HTTP methods are not exposed.
Authorization bypass in UVdesk's saved-reply endpoints lets an authenticated agent holding only ROLE_AGENT enumerate saved reply identifiers and read template content reserved for support groups and teams they do not belong to. Exploitation requires a valid low-privilege agent session and network reachability to the loadTicketSavedReplies / getSavedReplyContent XHR endpoints, and yields read-only exposure of saved-reply text with no integrity or availability impact; UVdesk core-framework versions before 1.1.7 (bundled in community-skeleton before 1.1.8) are affected. No confirmed active exploitation (no CISA KEV listing) and no public exploit code identified at time of analysis - this is a genuine but modest-severity information-disclosure flaw rather than an urgent priority.
MidnightBSD's mport package manager before 2.7.8 trusts the TMPDIR environment variable when extracting package metafiles, so an attacker who can influence the environment of a privileged (root or setuid/setgid) mport invocation can redirect temporary metadata extraction into a location they control. Exploitation requires a local foothold on a MidnightBSD host, a privileged mport run that inherits the attacker-supplied environment, and success in the extraction timing window, which keeps this a low-to-moderate-risk local path-integrity issue rather than a remotely exploitable flaw; no public exploit code or confirmed active exploitation (CISA KEV) was identified at time of analysis. The issue is fixed in mport 2.7.8 via upstream PR 123 (commit 3790fa4), which ignores unsafe TMPDIR values in privileged contexts and rejects empty TMPDIR.
A crafted bulk-insert payload triggers an integer overflow in FalkorDB's _BulkInsert_ReadProperty parser (/bulk_insert.c), crashing the module and taking down the Redis process that hosts it, for a denial-of-service condition in v4.20.1. The assessed CVSS 3.1 vector (5.3, AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L) means remote, unauthenticated, no-interaction exploitation with availability-only impact; in practice the effective bar depends on whether the Redis deployment enforces requirepass or ACL authentication, which would raise the requirement to valid credentials. No public exploit code identified at time of analysis, no CISA KEV listing, and the sole reference is the vendor GitHub issue FalkorDB/FalkorDB#2326 - consistent with an independent assessment of a moderate-severity, availability-only issue of low real-world priority.
Unauthenticated price tampering in the RestroPress WordPress plugin (versions before 3.4.6) lets any remote attacker alter the client-supplied add-on price when items are added to or updated in the cart, then place an order at an arbitrary - including zero - total. The flaw is a server-side validation failure affecting only sites that run RestroPress with online ordering/cart-and-checkout enabled and paid item add-ons configured; no credentials or user interaction are required and default configurations of the affected versions are exploitable. Impact is limited to integrity of pricing/billing (CVSS 5.3, C:N/I:L/A:N) rather than data disclosure or code execution, and no public exploit code or confirmed active exploitation has been identified at time of analysis.
Denial of service in the Zephyr ESP-hosted Wi-Fi driver (drivers/wifi/esp_hosted/) lets a Wi-Fi-adjacent, unauthenticated attacker fault the device or permanently stop Wi-Fi reception in Zephyr builds (assessed range 4.2.0 through 4.4.1) that run an ESP co-processor over SPI with CONFIG_NANOPB_ENABLE_MALLOC selected. The driver takes a 16-bit TLV data_length straight off the wire and passes it to pb_istream_from_buffer() without validating it against the frame length, so pb_decode() can read roughly 62 KB past a 3188-byte stack object (CWE-125); the underlying half-duplex SPI aliasing bug lets the attacker steer the parser onto misaligned protobuf continuation bytes by getting the first fragment of a fragmented control response dropped mid-transmit, and the AP scan list the attacker can grow is the most convenient lever. Impact is limited to availability (CVSS:3.1/AV:A/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:H, 5.3) with no disclosure or code execution; the high attack complexity reflects the timing race, no public exploit code identified at time of analysis, and the vendor has published a fix.
Unauthenticated SQL injection in Pagekit CMS through version 1.0.18 lets a remote attacker inject SQL through the credentials array posted to the public login endpoint (POST /user/authenticate), enabling read-oriented extraction of database contents such as the user table and password hashes. No authentication, user interaction, or non-default configuration is required, so the only meaningful limiter is exposure - any internet-reachable vulnerable instance is directly exploitable. Publicly available exploit code exists (a GitHub repository demonstrating the Pagekit login SQLi is referenced), but there is no CISA KEV entry and EPSS is just 0.15% (5th percentile); note the supplied base score of 5.3 understates impact, since our independent assessment rates confidentiality as High (CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N).
Incorrect access control in Stencil core 4.43.5 lets unauthenticated network attackers read a limited amount of information from the affected surface, consistent with the CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N vector where only confidentiality is impacted. There is no integrity or availability effect, no public exploit identified at time of analysis, and no CISA KEV listing; EPSS is 0.14% (4th percentile), placing it in the lowest risk band. Because Stencil is primarily a build and development toolchain rather than a production runtime, the vulnerable surface - most plausibly a dev server or generated assets - would normally need to be deliberately exposed on the network, which keeps real-world priority low despite the medium 5.3 base score.
WooCommerce PayPal Gateway before 9.2.1 allows unauthenticated attackers to mark their own orders as paid without paying the store by submitting a genuine PayPal transaction generated from a sandbox account they control. The plugin fails to verify that the payment notification came from the store's configured live payment environment and was paid to the store's merchant account, a business-logic flaw rated CVSS 5.3 (integrity impact only). No public exploit code or active exploitation has been identified, and stores running 9.2.1 or later are unaffected.
Concurrent writes to the same Internal Trusted Storage (ITS) UID in Zephyr RTOS can be encrypted with a duplicate AEAD nonce, because the default nonce provider secure_storage_its_transform_aead_get_nonce() keeps its counter in unsynchronized function-local statics (no lock). Affected are builds using the PSA ITS secure-storage transform with AES-GCM or ChaCha20-Poly1305, an application that writes the same UID from two threads at once, and one of the shipped log-structured back-ends (zms.c or the settings/NVS store) whose deferred garbage collection leaves both racing ciphertexts resident in the raw partition. The impact is cryptographic rather than direct code execution: nonce reuse under the same derived key leaks the XOR of the two plaintexts (potentially PSA persistent keys stored in ITS) and, for GCM, exposes the authentication key, allowing forgery of stored entries - but only for an adversary with local read access to the raw backing flash, and only where the same UID is written concurrently. CVSS is 5.3 (AV:L/AC:H/PR:L/UI:N/S:U/C:H/I:L/A:N), no CISA KEV listing, and no public exploit identified at time of analysis; risk is heavily gated by the narrow race window and the per-UID key salting that makes cross-UID collisions harmless. An upstream fix commit that serializes the provider with a mutex is available.
Memory exhaustion in the Checkmk agent receiver lets an attacker who controls a host registered for push mode send a tiny zlib-compressed payload that expands to an arbitrary size, consuming receiver memory and disrupting monitoring. Affected releases are Checkmk before 2.5.0p14, before 2.4.0p37, and before 2.3.0p51, plus the end-of-life 2.2.0 branch. No public exploit identified at time of analysis, and exploitation is gated by an established, authorized push-mode registration (authenticated per the PR:L vector), so realistic attackers are insiders or already-compromised monitored hosts; impact is limited to availability of the agent receiver with no data exposure or code execution.
Reflected cross-site scripting (XSS) in CuteNews 2.1.2 allows an unauthenticated remote attacker to inject a javascript: URI via the unsanitized __referer parameter, which is then rendered as a clickable link on the msg_info page. Exploitation requires an authenticated user (e.g., admin or editor) to view the msg_info page and click the crafted link, resulting in arbitrary JavaScript execution within that user's browser session. This is a low-priority issue with no public exploit code identified at time of analysis; CISA SSVC rates exploitation as none, automatable as no, and technical impact as partial. No vendor-released patch is available.
Reflected cross-site scripting (CWE-79) in Tautulli, the Python-based monitoring and tracking tool for Plex Media Server, affects all releases prior to 2.17.2 through the /search endpoint, which weaves the user-supplied query parameter into a JavaScript string. An unauthenticated attacker can craft a link that, once followed by an already-authenticated Tautulli user, executes arbitrary script inside that user's browser session on the Tautulli origin (CVSS 3.1: AV:N/AC:L/PR:N/UI:R/S:C/C:L/I:L/A:N). The flaw requires victim interaction and is confined to the Tautulli web context; no public exploit code has been identified at time of analysis, and the issue is fixed in 2.17.2.
Self-limited cross-site scripting in the 1millionbot AI Chat Platform (SaaS) enables unauthenticated remote attackers to inject Markdown-based external hyperlinks into their own interactive session. The flaw requires user interaction to render the unsanitised content and is confined to the attacker's own chat widget - it does not compromise other users, administrative panels, or backend infrastructure. No public exploit code or active exploitation (CISA KEV) has been identified at time of analysis, and the assessed real-world risk is low despite the network attack vector.
GoCD servers running versions 12.3.1 through 26.1.0 leak historical full server configuration through legacy /go/admin/restful/* routes to authenticated pipeline group administrators, who should only ever receive configuration for the pipeline groups they administer (CWE-863, incorrect authorization). Exposed data can include agent auto-registration keys, webhook invocation keys, encrypted material credentials and GoCD administrator lists, and a malicious group administrator could reuse a disclosed agent registration key to connect a rogue compatible agent, opening a higher-complexity path to receiving work or overwriting artifacts belonging to other groups. The endpoint is read-only and does not modify server configuration, normal authenticated users are explicitly not affected, and deployments that restrict pipeline editing to full administrators or to configuration repositories are not exposed; no public exploit identified at time of analysis. The issue is fixed in 26.1.0.
Stored cross-site scripting in Aureus ERP versions before 1.6.0 allows an authenticated user who can edit at least one tracked text field to persist malicious markup inside the Chatter field-change log, which executes in the browser of any other user - including an administrator - who opens the affected record's Chatter panel. The vulnerability is rooted in Laravel Blade templates that render the change log's old_value and new_value entries with the unescaped {!! !!} directive rather than escaped {{ }} output, and it has been confirmed at the source-code level by the vendor's fix commit. Vendor-released patch: 1.6.0; no public exploit identified at time of analysis, though a public technical writeup of the issue exists. Risk is moderate (CVSS 4.0 base 5.1; independent assessment CVSS 3.1 5.1 with S:C) because exploitation requires prior authentication plus victim interaction, but the stored nature and administrator-targeting potential make it a credible privilege-escalation vector inside multi-user deployments.
Stored cross-site scripting in UVdesk core-framework before 1.1.7 (and community-skeleton before 1.1.8) allows an authenticated agent to inject malicious script into the SwiftMailer configuration identifier field, which is persisted and later executed in the browser of any member who views the SwiftMailer configuration update page. The vulnerability is second-order and not self-contained, requiring a privileged victim to navigate to the affected page, and patched versions reject non-alphanumeric identifiers. No public exploit code has been identified at time of analysis, and the issue is assessed as low-to-moderate priority due to the agent-level authentication and user-interaction prerequisites.
Stored cross-site scripting in the Ditty News Ticker WordPress plugin (ditty-news-ticker) before 3.1.70 allows an authenticated user holding Author-level privileges or higher to inject arbitrary HTML elements into a layout, which then execute in the browser of any visitor who loads a post or page embedding the affected Ditty item. The flaw stems from ditty_layout_render_tag_wrapper() emitting caller-supplied wrapper attribute values directly as HTML element names without allowlist validation, so the payload survives because wp_kses_post() sanitization runs at save time, before those layout tag attributes are resolved at render time. There is no public exploit identified at time of analysis, and no evidence of confirmed active exploitation (CISA KEV); the risk is moderate and gated by the site's role-assignment policy, making it most dangerous on sites that grant Author or Editor access to untrusted or numerous contributors.
Missing object-level authorization in MaxKB 2.10.2-lts and earlier lets an authenticated, non-privileged user of a workspace read another user's hidden MCP tool record - including the Tool.code field, which can embed MCP server configuration and headers - by requesting the tool-detail route with a known tool_id. The same weakness permits a secondary abuse path: a user who can create or edit their own workflow can reference the foreign mcp_tool_id in a workflow MCP node so that workflow debug executes with the victim's MCP configuration without any permission check. The flaw is confined to members of the same workspace and requires prior knowledge of a valid tool_id, and no public exploit identified at time of analysis; no fixed version is available as of this review.
GoCD continuous delivery servers running 16.1.0 through 26.0.x allow a pipeline-group administrator to test SCM material connections belonging to pipeline groups and configuration repositories outside their authorized scope, causing disclosure of the credentials attached to those materials. Because the affected endpoints (/go/api/admin/internal/material_test and /go/api/internal/config_repos/*/material_test) accept an arbitrary pipeline and pipeline-group context, an attacker can also trigger resolution of external secrets managed by secret-management plugins on GoCD 20.6.0 and later, including global reusable secrets that carry separate permissions. Exploitation requires authenticated access with pipeline-group (or config-repo) administrator privileges (CVSS:3.1/AV:N/AC:L/PR:H/UI:N/S:U/C:H/I:N/A:N, 4.9), so this is a moderate-severity broken-authorization flaw rather than an emergency; no public exploit code was identified at time of analysis, and the issue is fixed in GoCD 26.1.0.
Identity-header forwarding in the Warpgate SSH/HTTPS/MySQL bastion host prior to version 0.25.6 lets an authenticated user be authorized as a different user by proxied HTTP backends that trust the first occurrence of the reserved x-warpgate-username header. The flaw is conditional rather than a drive-by critical: it requires (1) a valid authenticated Warpgate session (PR:L) and (2) a downstream backend that both trusts Warpgate's reserved identity headers and resolves duplicate header values by taking the first, since Warpgate forwards the client-supplied value ahead of its own injected authenticated username; backends that read the last value, deduplicate, or ignore x-warpgate-* headers are unaffected. It is rated CVSS 4.9 (AV:N/AC:H/PR:L/UI:N/S:C/C:L/I:L/A:N), has no CISA KEV listing, and no public exploit identified at time of analysis.
Missing authorization checks in the Keycloak Authorization Services policy evaluation endpoint allow an authenticated delegated administrator with limited viewing privileges to read the full profile and role information of any user in the realm, including email addresses and assigned security roles. This moderate-severity information-disclosure issue affects Red Hat Build of Keycloak and Red Hat Single Sign-On 7. No public exploit code has been identified at time of analysis.
Stale session cookies keep working in x-ui 0.3.2, so an attacker who already holds an administrator cookie issued before a credential change retains full access to the management panel even after the username or password has been rotated. The root cause is CWE-613 (insufficient session expiration): the complete user object is stored in a client-side signed cookie and authentication only verifies that a user object can be deserialized from it, with no re-validation against the database and no session-version or revocation check. Exploitation requires prior possession of a valid admin cookie (the assessed CVSS vector is CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:L/A:N, score 4.8, and does not permit a fresh unauthenticated login without a cookie); EPSS is 0.19% (8th percentile), there is no CISA KEV entry, no public exploit identified at time of analysis, and no vendor-released patch identified at time of analysis, so credential rotation alone cannot be relied on as containment.
Stored cross-site scripting in Tautulli versions prior to 2.17.2 allows a caller who already holds Tautulli administrator privileges or a valid Tautulli API key to persist crafted JavaScript inside the newsletter 'cron' field, which is then executed in the browser of an administrator who simply opens the newsletter configuration modal. The payload is stored in the newsletters table and keeps firing after API-key or credential rotation until the affected newsletter entry is manually deleted, giving the flaw a persistence property that outlives routine remediation. Exploitation requires high privilege to plant the value and a separate administrator interaction to trigger it, so no public exploit code has been identified and the practical risk is low; both the upstream advisory and the vendor release v2.17.2 confirm this is a genuine but low-urgency issue rather than a remotely exploitable, unauthenticated attack.
Stored cross-site scripting in Joplin (including Joplin Server deployments) before 3.7.2 lets a low-privileged authenticated Server user publish a crafted HTML note as a public share whose javascript: link passes the renderer's internal-resource validation. Exploitation is heavily constrained: the victim must middle-click or use 'Open in new tab' on the link and be running an older or non-hardened browser, because current Chrome and Firefox block javascript: new-tab navigation, and ordinary left-click is already blocked; if execution succeeds while the victim is signed in, script runs in the Joplin Server origin and can read page-visible content and issue authenticated same-origin requests. Rated CVSS 4.4 (AV:N/AC:H/PR:L/UI:R/S:C/C:L/I:L/A:N), no public exploit code identified at time of analysis and no confirmed active exploitation (not in CISA KEV); a vendor-released patch (3.7.2) is available.
Argument injection in Proxmox Mail Gateway's pmg-api package changelog retrieval functionality lets an attacker with an authenticated high-privileged session, or an attacker who can induce a logged-in PMG administrator to make a CSRF-style request, supply malicious arguments to the underlying apt-get command. The issue is classified as CWE-88 and tracked at CVSS 4.4 with high attack complexity, required administrative privileges, low confidentiality and integrity impact, and a scope change; no public exploit identified at time of analysis. Exploitation is limited to Proxmox Mail Gateway deployments and, in the CSRF variant, requires administrator interaction, making real-world risk low-to-moderate despite the network attack vector.
InvenTree versions prior to 1.4.0 expose business data through the report and label print endpoints because POST /api/report/print/ and POST /api/report/label/print/ authenticate the caller but never invoke users.permissions.check_user_permission against the target model's view role. Any authenticated account - including one with no roles or only unrelated roles - can enumerate object primary keys and render them through the default templates InvenTree ships for major business models, then retrieve the resulting artifact from DataOutput, obtaining purchase, sales, part, bill-of-materials, stock, build, supplier, customer, pricing, and inventory data that the corresponding detail APIs would correctly deny. Fixed in 1.4.0; no public exploit code or confirmed active exploitation was identified at time of analysis, and the EPSS-style risk is modest because valid credentials are required and impact is limited to confidentiality (CVSS 4.3, CWE-639).