An insecure direct object reference (IDOR) in ChatOllama (sugarforever/chat-ollama) through commit 96c3ce5 lets remote attackers read other users' private knowledge bases by supplying sequential integer identifiers. According to the supplied CVSS 4.0 vector (AV:N/AC:L/AT:N/PR:N/UI:N, VC:H with no integrity or availability impact, base 8.7), the exposure is a network-reachable, low-complexity, read-only disclosure that requires no privileges or user interaction, although defenders should confirm whether their deployment actually enforces authentication in front of the API. Affected endpoints are GET /api/knowledgebases/{id} and the knowledgebaseId parameter of POST /api/models/chat; the impact is limited to metadata and retrieved document content of other tenants, with no write or denial-of-service component. No public exploit code or confirmed active exploitation was identified at time of analysis; the primary risk driver is the predictable, non-tenant-scoped ID space rather than a remotely weaponized exploit.
Unauthenticated SQL injection in Thecosy IceCMS through 4.1.5 allows remote attackers to read arbitrary database contents by appending crafted SQL after the LIMIT clause that is concatenated into a MyBatis-Plus QueryWrapper for the num path parameter of /WebArticle/findarticlesbynum. Exploitation requires no authentication, no user interaction, and no special conditions beyond network reachability of the endpoint (CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N, base 7.5); the supplied CVSS 4.0 vector scores 8.7 with the same confidentiality-only impact. There is no CISA KEV listing and no confirmed public exploit code at time of analysis, though a public technical write-up describing the injection point is referenced, so exposure should be treated as a genuine high-priority data-disclosure risk for internet-facing instances that have not moved past 4.1.5.
Privilege escalation in EFAK (Eagle for Apache Kafka) through 5.1.0 lets any authenticated low-privilege user grant themselves ROLE_ADMIN, create new administrator accounts, or delete existing users by sending POST, PUT, or DELETE requests to the /api/users endpoints, which are handled by UserController but never restricted to administrators by WebSecurityConfig. Exploitation is not pre-authentication and requires a valid low-privilege EFAK account (CVSS PR:L), but there is no user-interaction requirement and standard configurations are affected; no public exploit identified at time of analysis. Success yields full administrative control over the Kafka clusters that EFAK manages, with high confidentiality, integrity, and availability impact, and urgency rises sharply in deployments allowing self-registration or shared low-privilege logins.
Privilege escalation in FastapiAdmin through version 3.0 allows any remote attacker who can reach the API to self-register a low-privileged account and then forge an HS256 JWT that impersonates the superuser, gaining full administrative control of the application. The flaw stems from a hard-coded JWT signing secret shipped in backend/app/config/setting.py (line 67), which lets an attacker sign arbitrary tokens while retaining their own valid session_id and simply setting user_name to the superuser; token validation in backend/app/core/dependencies.py (lines 72-112) trusts these forged claims. No user interaction, administrative access, or non-default configuration is required - only the ability to reach the registration endpoint and the API, which are present by default - and no public exploit code was identified at time of analysis, though a public technical writeup of the technique exists.
Credential disclosure in netdisk-fast-download through 0.4.8 allows unauthenticated remote attackers to retrieve donated netdisk credentials via the /v2/randomAuth endpoint and decrypt them using the hard-coded AES key 'nfd_auth_key2026'. This can expose cloud-drive cookies, tokens, or passwords, resulting in high confidentiality impact. Exploitation requires that the target instance has the donation feature enabled and populated with credentials; no public exploit has been identified at time of analysis.
Remote unauthenticated requests to a network-reachable CSAF Provider instance can exhaust CPU resources through unbounded resource allocation, leading to a denial of service. The flaw (CWE-770) affects the gocsaf development community's CSAF Provider, with no specific version constraints identified in the CPE data (all versions appear affected). No public exploit code was identified at time of analysis and there is no CISA KEV entry, though the CVSS 4.0 score of 8.7 reflects high availability impact (VA:H) with no confidentiality or integrity loss.
Account takeover of Google-linked bbs-go users is possible because the /api/login/google_one_tap endpoint accepts any Google-signed ID token without ever validating its audience (aud) claim, so the login handler trusts tokens minted for OAuth clients other than the site's own. An attacker who operates their own Google OAuth client can get a victim to sign in at an attacker-controlled property, capture the resulting Google-signed ID token, and replay it against bbs-go to obtain an authenticated session tied to the victim's Google-linked account (CWE-287). The flaw affects bbs-go through version 4.4.6 on instances where Google One Tap login is enabled and the target account is Google-linked; no public exploit identified at time of analysis, although a public technical writeup of the audience-confusion technique is referenced, and exploitation requires victim interaction to obtain a valid token.
SQL injection in Helical Insight through 7.0.0 allows authenticated users with at least viewer privileges to execute arbitrary SQL queries via crafted Numeric or String placeholder values sent to the /executeDatasource endpoint, bypassing the broken SqlDetect guard. This enables reading unexposed database tables and modifying data within the report's configured data source, with impact limited by the privileges of the database account used by the application. No public exploit code or active exploitation has been identified at time of analysis.
Remote, unauthenticated attackers can access password-less development tools under /wheels/* in Wheels 4.0.0 through 4.1.2 when enablePublicComponent=true (the default in development mode), exposing application settings, environment and datasource details, and routes, and allowing them to start test runs that may write to the real application database. This incorrect authorization flaw (CWE-863) does not affect production-mode deployments by default, and the separately protected migrator and CLI endpoints are not impacted. No public exploit identified at time of analysis.
Cross-workspace authorization bypass in Jpom through 2.11.12 (CWE-639, IDOR) lets any authenticated user who holds SSH file permissions in one workspace read, modify, or delete files on SSH hosts that belong to a different workspace. The flaw resides in SshFileController, where the caller-supplied SSH asset id is used to resolve the target host without verifying that the asset actually belongs to the caller's workspace, so the reachable endpoints (download, upload, read_file_data.json, update_file_data.json, delete.json, rename.json, new_file_folder.json) operate on foreign assets. Exploitation requires a multi-workspace deployment with at least one SSH host asset outside the attacker's workspace and knowledge or enumeration of that host's asset id; no public exploit code is confirmed at time of analysis, though a public technical write-up (linked from the CVE record) explains the technique, and the issue is not listed in CISA KEV, so there is no evidence of active exploitation. Assigned CVSS 4.0 base 8.6 (with our 3.1 assessment at AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:N) reflects high confidentiality and integrity impact while the CVSS 4.0 vector shows no availability impact.
Hard-coded cryptographic key material embedded in the /usr/bin/sysman binary of Moxa's AIG-302 Series industrial computers enables an unauthenticated attacker who already sits on the same local or logical network segment to recover the static private key and use it to decrypt or tamper with the device's protected communications. The key is identical across all affected units and is trivially extractable by reverse-engineering shipped firmware, so the confidentiality and integrity of those communications are effectively compromised once an attacker reaches the industrial LAN; availability is unaffected and no public exploit code has been identified at time of analysis, with no confirmed active exploitation. The practical barrier is the required adjacent-network position rather than any credential or user interaction, making this a moderately high priority OT/ICS exposure rather than an internet-reachable mass-exploitation issue.
Broken access control in LingGuo-Drama at and through commit 8a6924adf8142678919b4d9a08b9d60c49dcfbed lets any self-registered, low-privilege user read, overwrite, reassign, or delete other users' short-drama projects by iterating numeric IDs against the admin API endpoints GET, PUT, and DELETE /admin/v1/projects/:id. The root cause is CWE-639: the projects.Get helper omits an owner check, so the object identifier supplied by the caller is trusted without verifying that the authenticated principal owns the referenced project. Exploitation requires an authenticated account (CVSS PR:L, obtainable through self-registration where enabled) and network reach to the admin API; no public exploit identified at time of analysis, and there is no KEV entry or EPSS signal to suggest active exploitation in the wild.
An authorization bypass in HestiaCP before 1.10.5 lets an authenticated administrator who is not the protected ROOT_USER impersonate that account through the built-in 'loginas' parameter and then submit write requests, such as a POST to /add/cron/, that the application erroneously executes because the read-only flag is never validated. The practical result is command execution as the protected administrator's system account, since crond runs the injected job, giving high confidentiality, integrity, and availability impact. Exploitation requires a valid admin-level account on an installation with more than one admin and a protected ROOT_USER to shield; the vector is CVSS 4.0 AV:N/AC:L/PR:H/UI:N (vendor 8.6) and the independent assessment concurs, with patch available. No public exploit code or confirmed active exploitation was identified at time of analysis.
Authenticated account takeover in Huly Platform through 0.7.432: the mergeSpecifiedPersons account RPC lacks any authorization or workspace-role check, so any logged-in user - not just a workspace Maintainer - can merge arbitrary person records. An attacker who knows the victim's person UUID and an attacker-controlled person UUID can graft an attacker-owned, unverified but login-capable social id onto the victim person and then complete the takeover through the normal login or password-recovery flow. Vendor patch is available (upstream commit afe11faec1b56f278c0882355a69b08cff1eebd9); no public exploit identified at time of analysis, and real-world risk is bounded by the PR:L prerequisite plus the need to obtain two UUIDs, making instances with open self-service registration the highest-exposure targets.
Improper validation of the reset_url parameter in the Checkmk Setup in-page search lets an unprivileged remote attacker craft a URL that, once opened by an authenticated user holding Setup access, executes arbitrary script in that user's session or redirects the user to an arbitrary external site - but only when the victim clicks the Reset button or submits an empty search, since the payload does not fire on page load. Affected are Checkmk releases before 2.5.0p15, before 2.4.0p38, before 2.3.0p51, and the end-of-life 2.2.0 branch; fully patched instances are not exploitable. No public exploit code has been identified at time of analysis, and exploitation depends on successful social engineering of a privileged Setup user combined with mandatory user interaction.
Arcane versions prior to 2.5.0 disclose stored container-registry credentials to an attacker-controlled host: an authenticated principal holding both the registries:update and registries:test permissions can repoint an existing generic (non-ECR) registry entry to a URL they control without supplying a replacement token, causing the preserved encrypted PAT or password to be decrypted and sent to that host during RegistryLogin/TestRegistry. The issue is rated CVSS 8.5 (AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:L/A:N) and classified as CWE-200 information exposure, with the scope change and high confidentiality impact reflecting that the leaked credential grants access to, and potential tampering with, the source registry's images. No public exploit code and no confirmed active exploitation were identified at time of analysis, but the flaw is a credible credential-theft primitive for any tenant or operator account that has been granted registry management rights. A vendor fix is available in the 2.5.0 release.
TuxGuitar versions before 2.1.0 will load and instantiate an attacker-supplied Java class contained in a crafted .tg tablature file, executing that class's constructor (and static initializer) with the privileges of the TuxGuitar process when the user opens the file. The document sets the gervil.soundbank.path property to an on-disk archive - the .tg file itself qualifies - whose META-INF/services/javax.sound.midi.Soundbank entry names a class bundled in that same archive, and the bundled Gervill media.sound.JARSoundbankReader loads it through an unverified URLClassLoader. The CVSS 4.0 vector is AV:L/AC:L/AT:N/PR:N/UI:P with a score of 8.5 (unauthenticated, no non-default configuration required), but the payload must be delivered as a document the victim opens, so real-world priority is moderated by that delivery model; no public exploit code has been identified at time of analysis and the issue is fixed in 2.1.0.
SQL injection in the MapSVG WordPress plugin (mapsvg-lite-interactive-vector-maps) affects all versions through 8.16.1 and lets an authenticated low-privileged user, such as a subscriber, contributor or customer account, inject arbitrary SQL through the plugin's vulnerable request handling, exposing database contents. The issue is remotely reachable with low attack complexity and needs no victim interaction, but it does require a valid account on a site that both runs the affected plugin and permits untrusted low-privilege registration; sites that do not hand out such accounts face materially lower real-world exposure. No public exploit code has been identified at time of analysis and the flaw is not listed in CISA KEV, so the practical risk is driven by the population of sites exposing low-privileged authenticated access to the plugin.
Authenticated non-admin users of sshportal's management shell can retrieve decrypted backend SSH private keys through the 'key show' command, allowing them to connect directly to managed hosts and bypass ACLs and session recording. The flaw affects sshportal through version 1.19.5 and requires only a valid non-admin account with management shell access; default deployments are impacted because predictable key names such as 'default' and 'host' are commonly present. No public exploit code or confirmed active exploitation was identified at time of analysis, but the CVSS 4.0 score is 8.3 (High) and our assessment rates it as a medium-to-high priority for shared multi-user bastion deployments.
SQL injection in Frappe HRMS 14.x (before 14.38.2), 15.x (before 15.64.0), and 16.x (before 16.18.0) allows any authenticated user, including a low-privilege account, to extract sensitive information from the database via crafted requests to endpoints that lack input validation. The flaw requires only a valid session (CVSS PR:L) and no user interaction, resulting in high confidentiality impact with scope change. Patched versions 14.38.2, 15.64.0, and 16.18.0 are available, and no public exploit code has been identified at time of analysis.
The Salon Booking System WordPress plugin before 10.31.3 fails to enforce capability checks on several of its booking REST API routes, so any authenticated user - including a low-privilege subscriber - can retrieve the booking records and personal data of every customer on the site. The vulnerability is remotely reachable (CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:N/A:N, base score 4.3) but requires a valid account, meaning sites that allow self-service registration or that host untrusted low-privilege users are the realistic attack surface. Publicly available exploit code exists via the WPScan disclosure, though EPSS is low (0.15%, 4th percentile) and no active exploitation has been confirmed. This is a genuine but moderate-priority information disclosure issue rather than a top-tier emergency.
The Visual Form Builder WordPress plugin through version 3.1 renders a wp-admin dashboard widget that lists recent form submissions without performing any capability or authorization check, so any authenticated user at Subscriber level or above can read other users' submitted contact information - names, email addresses, form titles and timestamps. Exploitation requires a valid low-privilege WordPress account (CVSS PR:L) and the ability to reach the dashboard where the widget loads, and the impact is confined to read-only disclosure with no data modification or account takeover. Publicly available exploit code exists, but there is no confirmed active exploitation, no CISA KEV listing, and the EPSS score is negligible at 0.15% (4th percentile), so this is a genuinely low-priority issue that is nonetheless a real and easily fixed flaw.
Authenticated users with Subscriber-level access can persistently corrupt the field structure of any form built with Visual Form Builder through 3.1, because the plugin's field-reordering action performs neither an authorization nor a nonce check (CWE-862). Publicly available exploit code exists, but EPSS is just 0.14% (3rd percentile) and CISA KEV does not list this issue, so active exploitation is not confirmed. The impact is limited to integrity-forms break for public visitors-with no data disclosure, and exposure is greatest on sites that permit open user registration.
Missing authentication in the OtterMind Nubase storage-object API (versions through 0.1.4) lets remote attackers download and delete objects held in private buckets without supplying an apikey or JWT. The UnifiedMultiTenancyFilter treats the route /storage/v1/object/{bucketId}/{path} as public for every HTTP method, so any request reaching that endpoint bypasses tenant authorization entirely; per the independent assessment the impact is a high-confidentiality, low-integrity/availability compromise (CVSS:3.1 8.3-equiv, AV:N/AC:H/PR:N/UI:N/C:H/I:L/A:L), gated only by the attacker's prior knowledge of a valid tenant subdomain, bucket id and object path. No public exploit code or CISA KEV entry was identified at time of analysis, and no vendor-released patch is confirmed.
Cross-domain sender spoofing in Mailu allows a domain manager to authenticate as arbitrary senders on other domains hosted on the same instance. The flaw is an improper-authorization issue in which the Postfix sender-login lookup honors the allow_spoofing flag globally rather than scoping it to the requesting user's own domain, so a manager of one hosted domain can enable spoofing and pass DKIM/SPF authentication for domains they do not own. Exploitation requires an existing domain-manager (administrative) role (CVSS PR:H) on a multi-tenant deployment hosting more than one domain; there is no unauthenticated or end-user path, and no public exploit code was identified at time of analysis.
Arbitrary file deletion in the Groundhogg CRM, Newsletters and Marketing Automation plugin for WordPress (all versions up to and including 4.9.3) lets an authenticated user who holds the plugin's import_contacts capability - granted by default to Groundhogg's custom Sales Representative role - name an arbitrary server-side path and have the plugin delete it, with a realistic escalation to remote code execution (for example by removing wp-config.php). The flaw is reachable through the contact-import (CSV), REST API, and webhook paths that call copy_file, and it received a CVSS of 8.1 (AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:H/A:H); there is no CISA KEV listing and no public exploit code identified at time of analysis. The practical risk is bounded by the authentication requirement and by the attacker's need to know or guess the absolute path of a valuable target, but the capability is assigned by default and no user interaction or non-default configuration is needed, making this a credible, actionable priority rather than an inflated high-CVSS issue.
Unauthenticated remote attackers can retrieve full payment records from Roncoo-pay by exploiting a missing authorization check in the F2FPayController.orderQuery endpoint. The vulnerability (CWE-306) allows access to any merchant's transaction data if the attacker can guess or enumerate a valid transaction number, which follows a predictable format ('7' prefix plus a snowflake ID). No public exploit code is known at this time, and exploitation requires network access to the web gateway and successful guessing of a transaction number, making it a complexity-limited information-disclosure issue.
DNS cache poisoning in Amazon FreeRTOS-Plus-TCP prior to V4.4.2 lets unauthenticated remote attackers insert an attacker-chosen address into a device's DNS cache by supplying a forged DNS reply that the resolver accepts before the legitimate response arrives. The flaw is an integrity-only weakness (CWE-940, improper verification of the source of a communication channel): the resolver does not adequately validate that a reply actually originated from the DNS server it queried, so a successful injection diverts subsequent outbound connections to attacker-controlled endpoints. Exploitation is constrained by high attack complexity - the attacker must match or guess reply-matching attributes such as source address/port and transaction ID and win a timing race - which favors on-path or local-network attackers over arbitrary internet senders; no confirmed active exploitation (CISA KEV) and no public exploit code have been identified at time of analysis, and the vendor has released a fixed version with a required configuration flag.
Missing authorization and CSRF protections in the MB Custom Post Types & Custom Taxonomies WordPress plugin before 2.12.2 allow any authenticated user - including a low-privileged subscriber - to trigger the plugin's one-way migration routine, which creates unwanted published records and permanently deletes the site's entire post-type and taxonomy configuration. The migration actions perform neither a capability check nor nonce validation (CWE-862 broken access control), so no administrator role or valid token is required; an independent assessment rates the availability impact as High because the configuration loss is irreversible. Publicly available exploit code exists (publicly available exploit code exists), but EPSS is low at 0.14% (3rd percentile) and there is no confirmed active exploitation; the primary practical prerequisite is that the target site accepts self-service/low-privilege registration.
Privilege escalation in the Customer Email Verification for WooCommerce WordPress plugin through 3.3.0 allows unauthenticated attackers to gain authentication as any WordPress account that still has an unverified email address, including administrators, by brute-forcing a per-user activation secret that has only about 101 possible values. The weakness is genuine and high impact (full account takeover), but real-world risk is heavily gated by three preconditions: the plugin's 'Encoding method' must be set to the non-default 'hashids' option, the attacker must recover or brute-force the Hashids salt (which defaults to an md5(time()) value fixed at first initialization), and the targeted account's email must remain unverified. An upstream changeset addressing the issue is referenced in the WordPress plugin repository, and no public exploit code was identified at time of analysis; CVSS 3.1 scores it 8.1 (High, AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H), so despite the High score it should be triaged as lower priority than a default-configuration, unauthenticated account-takeover bug in most deployments.
Vulnity Security, a WordPress plugin maintained by Manuel Galan, can be bypassed to gain administrator access on Nginx-based installations running version 1.4.3 or earlier. The plugin's delegated-login feature writes a pairing credential file and auto-generates an Nginx configuration snippet intended to block web access to it, but the snippet guards the wrong directory path, leaving the credential file reachable by unauthenticated attackers who can then forge signed REST API requests and authenticate as an administrator. No public exploit code or confirmed active exploitation was identified at time of analysis, but the attack requires no authentication and no user interaction (CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H, 8.1), with High complexity arising from the dependency on a qualifying Nginx deployment where the delegated-login pairing credential exists and the flawed snippet is active; Apache and other web servers appear unaffected.
Unauthenticated stored-file disclosure in SurveyKing before 1.13.0 lets remote attackers download arbitrary attachments by calling GET /api/public/preview/{attachmentId}, bypassing the shared, owner, and project authorization checks that FileServiceImpl.loadFile is supposed to enforce. No authentication (PR:N), user interaction, or non-default configuration is required - the endpoint is public by design - but there is no file-listing capability, so success depends on knowing or deriving a valid snowflake attachment ID, which is time-ordered and only partially predictable (reflected in the assessed AC:H/AT:P). The result is a genuine but conditionally gated confidentiality risk rather than a mass-exploitation emergency: no confirmed active exploitation (not in CISA KEV) and no public exploit code identified at time of analysis, though a vendor patch exists and the exposure is high-severity (CVSS 4.0 8.2; CVSS 3.1 AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:N/A:N).
Denial of service in mz-automation libIEC61850 before 1.6.2 lets a remote, unauthenticated attacker crash IEC 61850 servers by sending a single defineNamedVariableList MMS request. The bug is a stack out-of-bounds read (CWE-125) in checkDataSetAccess, where source and destination arguments are reversed in a StringUtils_copyStringToBuffer call used to build an MMS_ASSOCIATION_SPECIFIC dataset reference, so an uninitialized 130-byte stack buffer is treated as a source string and strlen walks past its bounds before those bytes are copied into the caller-owned listName buffer. Exposure is conditional: the server must have the dynamic dataset service enabled and a dataset access handler installed, both non-default integration choices, but any network-reachable server configured that way is trivially hit for a full loss of availability; the vulnerability is fixed in release v1.6.2 and no public exploit code was identified at time of analysis. No CISA KEV entry is associated with this CVE, and the tags indicating information disclosure reflect leaked stack bytes rather than a confirmed confidentiality breach, since the assessed impact is availability-only.
Unauthenticated account takeover of WordPress sites running miniOrange OTP Login, Verification and SMS Notifications through 5.5.8 is possible because the Delay OTP grace flag is stored per user rather than being bound to the originating session, cookie, IP address, or device. When OTP-only login and Delay OTP Verification are enabled and the target account is inside its grace interval, an attacker can submit a known username with `mo_wp_login_intent=otp` and any non-empty password to inherit that user's verified state and log in as any account, including administrators. No public exploit code or CISA KEV active exploitation was identified at time of analysis, and exploitation is constrained by non-default settings plus a race-window timing precondition (CVSS 8.1, AC:H).
Unauthenticated privilege escalation in the Cozmoslabs User Profile Builder WordPress plugin (all versions up to and including 4.0.3) allows an attacker to obtain a fully activated administrator account: they submit role=administrator through the public [wppb-register] registration form, then click the activation link emailed to them, and the activation handler applies that role via WP_User::set_role() without any allow-list check. The attack is conditional rather than trivial - the site must already carry a legacy registration form field with meta-name 'role' from before the 4.0.3 upgrade (the 4.0.3 reserved-name block deliberately exempts pre-existing fields), must have email confirmation enabled so the wp_signups path is used, and must allow open user registration. That combination of stable site-state prerequisites is why CVSS 3.1 scores it 8.1 with AC:H despite PR:N/UI:N, and no public exploit code has been identified at the time of analysis.
Broken object-level authorization in the Wheels CFML framework (versions 1.1 through 4.1.2) allows an authenticated low-privileged user to tamper with other users' associated records through nested forms. The nestedProperties() helper resolves nested hasMany, hasOne, and belongsTo records by the primary key posted in the request without verifying that those rows belong to the parent record being saved, so an attacker who can legitimately submit a nested form for their own parent can reassign, edit, or - when allowDelete=true is configured - delete another parent's rows, and can also create new child rows with attacker-chosen identities when the child model uses a composite primary key that is not composed of foreign keys. The issue carries an assessed CVSS of 8.1 (AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:N), requires authentication and actual use of nestedProperties() with nested forms, and is fixed in Wheels 4.2.0; no public exploit identified at time of analysis.
Broken object-level authorization in the Wheels (CFML) framework 4.0.0 through 4.1.2 lets an authenticated low-privilege attacker supply a record key to findByKey(), exists(key=), updateByKey(), or deleteByKey() and have the framework drop the application's scope condition (such as a tenant or owner filter), enabling read, modification, or deletion of another user's or tenant's record. The flaw is remotely exploitable (AV:N, PR:L, no user interaction) with high confidentiality and integrity impact but no availability impact, and the exists(key=) variant additionally leaks record existence by returning true for an out-of-scope key when the application runs in production mode with error information disabled. No public exploit code or CISA KEV entry is identified at time of analysis, and exploitation requires an application that relies on Wheels scopes for ownership or tenant authorization and passes request-supplied keys into the affected methods.
PHP object injection in the Gravity Forms WordPress plugin (all versions before 3.1.1) allows unauthenticated, network-reachable attackers to feed crafted serialized data into a PHP unserialize() call, which can lead to object injection and, if a suitable gadget chain is present in the running PHP stack, to full compromise of confidentiality, integrity, and availability. The flaw is rated CVSS 8.1 with vector CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H, meaning no credentials and no user interaction are needed, but the high attack-complexity component reflects that exploitation depends on a usable POP gadget chain existing in the plugin, WordPress core, or another installed plugin or theme. At time of analysis there is no public exploit identified (no CISA KEV listing and no published proof-of-concept in the supplied intelligence), and the vulnerability was reported through Patchstack to the vendor, with version 3.1.1 identified as the fixed release.
Broken object-level authorization in Tandoor Recipes before 2.6.12 lets an authenticated member of the same Space, holding read and write API scope, read and silently alter another member's private recipe instructions and ingredients by supplying those private step or ingredient identifiers to the Step/Ingredient API. The affected sub-resource ViewSets (StepViewSet, IngredientViewSet) filter only on active Space and apply role-level CustomIsUser checks instead of the private-recipe owner/sharing predicate, so a co-tenant is accepted without recipe-level authorization. No public exploit code has been identified at time of analysis, and the flaw requires legitimate write-scoped credentials in the victim's Space; it is fixed in version 2.6.12.
Apache Answer through 2.0.2 writes full API keys into application logs on every authenticated request, so anyone who can read those logs - directly or through downstream SIEM, log aggregation, backup archives, or support bundles - can recover and replay still-valid bearer credentials. The authoritative assessment scores this locally (AV:L, PR:L) rather than over the network, because log-read access is the gating prerequisite; an attacker without that position gains nothing. Recovered tokens must remain unexpired and unrevoked to be replayed, instances on 2.0.3 or later are unaffected, and no public exploit code has been identified at time of analysis (EPSS 0.20%, no CISA KEV listing), which supports a moderate rather than urgent priority.
Session hijacking via cross-site request forgery affects OpenHuman versions before 0.56.0: the unauthenticated loopback callback GET /auth/telegram accepts an attacker-supplied one-time Telegram login token and writes the returned JWT straight into the victim's local session store, switching the victim's active session to the attacker's backend account (CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:N, base 8.1 High). Because the malicious request is issued by the victim's own browser to 127.0.0.1:7788, the attacker needs only their own OpenHuman account, a valid one-time Telegram token, and a victim who visits a crafted page while their local core is running - no authentication to the victim host is required. Exploitation leaves subsequent agent actions, memory writes, and Telegram messages flowing into the attacker's account; no public exploit code or confirmed active exploitation was identified at time of analysis.
Arbitrary file and directory deletion in Countly Server (versions prior to 24.05.50 and 25.03.44) allows an authenticated user holding data-migration permission to pass a crafted exportid to the migration-deletion route and recursively remove any file or directory tree writable by the Countly server process, resulting in data loss and service disruption. The root issue is a path-traversal flaw (CWE-22) in plugins/data_migration/api/api.js in which the user-controlled exportid reaches clean_up_data in data_migration_helper.js without the sanitizeFilename-based containment checks introduced by the fix. The assessed CVSS 3.1 vector is AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:H/A:H (base 8.1), confirming a network-reachable, low-complexity attack requiring an authenticated account (PR:L) with no user interaction, with high integrity and availability impact but no confidentiality exposure; no public exploit code was identified at time of analysis and there is no confirmed active exploitation (not in CISA KEV). Vendor fixes are available in 24.05.50 and 25.03.44.
Privilege escalation to full WordPress site takeover is possible in the Backuply backup/restore plugin (all versions up to and including 1.5.8) via a second-order SQL injection that an authenticated subscriber-level user plants in the profile Biographical Info field and that only fires when an administrator later creates a backup capturing that value and then restores it. During the restore the injected SQL executes blind inside an INSERT into wp_usermeta, allowing the attacker to alter database records - most notably to set the default_role option in wp_options to administrator and enable open registration, so a subsequently registered account is granted Administrator rights. Exploitation cannot be self-triggered by the attacker (UI:R) and requires a valid authenticated account (PR:L, no anonymous path); no public exploit identified at time of analysis.
PHP object injection in the Search Regex WordPress plugin version 3.5 lets an authenticated attacker with subscriber-level access or above pass crafted serialized data into the plugin's replace_all function, which is then deserialized by PHP without validation. When a usable POP gadget chain exists in the installed plugin, WordPress core, or another active plugin, this object injection can be escalated to arbitrary code execution, giving the attacker full control of the site. No public exploit code or confirmed in-the-wild exploitation was identified at the time of analysis, and exploitation is conditional on both an authenticated account and a viable gadget chain.
Command injection (CWE-78) and stored XSS in Wheels CLI 4.0.0 through 4.1.1 allow an attacker who can influence inputs processed by the tool to run arbitrary shell commands as the SSH deploy user on deploy hosts and on the CI runner during builds. Deploy release/version/rollback values, package removal/destroy and extraction paths, and generator names are the primary injection surfaces, with unencoded scaffold 'show' views (since 4.1.0) and generated admin views permitting stored XSS; trigger inputs can come from CI variables, Git branch names, downloaded packages, or an AI agent using the CLI's MCP tools. The flaw is fixed in 4.1.2, and existing affected generated views also require a documented manual edit; no public exploit code has been identified at time of analysis.
Local privilege escalation in cifs-utils lets an unprivileged local user obtain root on affected hosts by injecting unauthorized options into a privileged CIFS mount. When a root-owned service mounts a Common Internet File System share using a credentials file whose contents the attacker can control or influence - for example a file that is world-writable or lives in an attacker-owned path - cifs-utils assembles those unsanitized values into the mount option string, so the attacker can slip in extra options that bypass permission checks and cause files to be executed with administrative privileges. The flaw is tracked as CWE-88 (argument/option injection) with a CVSS of 7.8 (AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H); exploitation requires an existing low-privilege local account (authenticated, per PR:L) and a deployment where such a privileged, credentials-file-driven CIFS mount workflow exists, so internet-facing hosts that do not mount attacker-influenced CIFS shares are not at risk. No public exploit code or confirmed active exploitation was identified at time of analysis, and Red Hat has published advisory and Bugzilla tracking pages for RHEL 6 through 10 and OpenShift Container Platform 4.
Incorrect DBus policy in Lomiri Download Manager (lomiri-download-manager on UBports) allows any local user to read arbitrary files as root via the privileged DBus service, resulting in information disclosure. This local privilege escalation vulnerability affects all versions and requires local access to a device running UBports Lomiri/Ubuntu Touch with the service active. No public exploit code or active exploitation has been identified; risk is moderate due to the local-only attack vector.
Local privilege escalation in IBM i 7.3, 7.4, 7.5, and 7.6 allows an authenticated, low-privileged user to obtain QSECOFR/Root authority, as described in CVE-2026-102917. The vulnerability has a CVSS 3.1 base score of 7.8 (AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H) and requires prior authentication and local access; it is not remotely exploitable. No public exploit code has been identified at time of analysis, and IBM has released a patch via its support advisory.
Opening an attacker-crafted repository in Z.ai ZCode v3.11.2.6792 can lead to arbitrary code execution because the IDE automatically connects to MCP (Model Context Protocol) servers declared in the repository's workspace configuration without requiring user approval. The flaw is rated CVSS 7.8 with a local attack vector and required user interaction - a victim must actively open the malicious repository - and publicly available exploit code exists, though there is no confirmed active exploitation (not in CISA KEV), so the threat is conditional and interaction-dependent rather than a mass-exploitation emergency. Exploitation is unauthenticated from the attacker's perspective and grants high confidentiality, integrity, and availability impact on the affected host.
Improper quoting during SQL serialization in MariaDB Server 10.6.1 through 10.6.27, 10.11.18, 11.4.12, 11.8.8, 12.3.2 and 13.0.1 and earlier allows a low-privileged authenticated attacker who can create a foreign-data-wrapper SERVER, USER, UDF, plugin, or trigger with crafted identifiers or option values to inject arbitrary SQL into the output of SHOW CREATE SERVER, SHOW CREATE USER, or mariadb-dump; the injected statements execute only when a privileged operator later replays that generated SQL, such as during a dump restore. Exploitation requires both a pre-existing low-privileged account (CVSS PR:L) and a victim-initiated action (UI:R), so this is a conditional, stored-injection-style threat rather than an unauthenticated remote compromise. No public exploit code has been identified at time of analysis, and the issue is fixed in MariaDB 10.6.28, 10.11.19, 11.4.13, 11.8.9, 12.3.3 and 13.0.2.