OS command injection in the H3C Magic NX15 consumer router (confirmed on firmware 100R017) lets a remote attacker who already holds a low-privilege authenticated session run arbitrary operating-system commands on the device via the sub_416874 function of /usr/bin/usrlist. The flaw scores 8.6 under CVSS v4.0 and 8.6-equivalent High under the assessed CVSS 3.1 vector (AV:N/AC:L/PR:L/UI:N with full C/I/A impact), and publicly available exploit code exists - a working proof-of-concept and writeup for the usrlist command-injection path have been published on GitHub. The vulnerability is not listed in CISA KEV, so this is a public-PoC threat rather than a confirmed in-the-wild campaign; the practical risk hinges on whether the management interface is reachable from an untrusted network, since default deployments typically keep it LAN-side.
Path traversal in MLflow’s FileStore tracking backend allows unauthenticated remote attackers to write attacker-controlled trace metadata outside the intended experiment traces directory by submitting a crafted trace ID to the POST /api/3.0/mlflow/traces endpoint. MLflow 3.17.1 and earlier are affected; FileStore is the default backend before 3.7.0 and opt-in from 3.7.0 onward, so exposure depends on backend selection and network reachability. Publicly available exploit code exists, but no CISA KEV confirmation is present in the provided data; assessed impact is high integrity and low availability with no confidentiality impact.
Code injection in the H3C Magic NX15 home router (firmware NX15V100R017) allows a low-privilege authenticated attacker to manipulate the `path` argument of the `write` function exposed at /api/esps and achieve arbitrary file writes that escalate to root-level remote code execution. The flaw is remotely reachable over the RPC management interface (CVSS 4.0 score 8.6, AV:N/AC:L/PR:L/UI:N with high confidentiality, integrity, availability and subsequent-system impact; assessed as CVSS 3.1 AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H), with the PR:L prerequisite and the network exposure of the management interface acting as the main practical limits rather than attack complexity. Publicly available exploit code exists, including a published proof-of-concept for the file-write-to-root-RCE chain; no CISA KEV entry or confirmed in-the-wild exploitation was provided in the source data.
Missing authorization checks in Camunda 7's engine-rest task sub-resources allow any authenticated user to read, download, or permanently delete task comments and attachments without holding any permission grant on the target task. Versions 7.24.0 through 7.24.13 and 7.23.0 through 7.23.16 are affected; the attacker only needs a valid session and the target comment or attachment object ID. Publicly available exploit code exists and the issue is rated 8.6 under CVSS 4.0, making this a genuine priority for deployments that expose engine-rest.
Unauthenticated remote attackers can crash the SGLang inference server (versions up to and including 0.5.19) by calling its expert-distribution recording control endpoints - /start_expert_distribution_record, /stop_expert_distribution_record and /dump_expert_distribution_record - on a server that was launched without --expert-distribution-recorder-mode, which is the default configuration. Because the recorder is never initialized in that state, the request path into Scheduler.expert_distribution_handle fails without a handled error, producing a denial of service against the serving process; confidentiality and integrity are unaffected. Publicly available exploit code exists (a proof-of-concept is linked from the vendor issue tracker), but there is no evidence of confirmed active exploitation (no CISA KEV listing); the practical gate on real-world abuse is whether the SGLang control API is reachable from an untrusted network, since these endpoints are typically bound to internal or trusted segments. Vendor-released patch: not confirmed - the fix is an upstream pull request (sgl-project/sglang#42098) that is still awaiting acceptance.
Command injection in PingTools Network Utilities 4.64 Free allows a local user who can set or influence the Host field of a Watcher ICMP service to run arbitrary commands through sh -c, compromising confidentiality, integrity, and availability. The assessed vector is AV:L, so this is a local-device issue requiring access to the app and configuration or triggering of the Watcher ICMP feature, not an internet-facing remote attack. No public exploit code or CISA KEV entry was identified in the provided data.
Heap buffer overflow in IBM MQ Operator and IBM MQ Advanced Container Images lets an authenticated user with authority to perform MQPUT operations crash the queue manager or potentially execute arbitrary code by supplying malformed message options. The advisory is scoped to these containerized IBM MQ products, and exploitation requires valid MQ credentials plus put authority, so this is not an unauthenticated attack. No public exploit identified at time of analysis, and IBM has released a vendor patch.
Denial of service and potential remote code execution in IBM MQ deployments using the MQ Operator and MQ Advanced Container Images result from a heap buffer overflow when an authenticated attacker with put authority sends MQPUT or MQPUT1 operations carrying malformed message handle properties. The CVSS 8.8 vector is network-reachable and requires no user interaction, but the PR:L component means the attacker must already hold valid credentials and put permissions, so the practical threat is an insider or post-compromise actor rather than an unauthenticated remote attacker. No public exploit code was identified at time of analysis, and IBM has released a vendor patch.
Out-of-bounds read in Pillow's TIFF encoder before 13.0.0 allows unauthenticated remote attackers to leak heap memory or cause a denial of service (process crash) when an application saves or re-encodes a malicious TIFF using the libtiff encoder. Exploitation requires the target to re-encode the attacker-supplied image-simply opening it does not trigger the bug-and the leak is limited to adjacent heap bytes rather than targeted data. No public exploit has been identified at time of analysis, and the vendor has released Pillow 13.0.0 to fix the issue.
Heap out-of-bounds over-read in Pillow 8.1.0 through 12.x allows a crafted TIFF processed by an application using libtiff-backed decoding to copy adjacent heap contents into output pixels or crash the process. The flaw affects the tile path in _decodeTile and the RGBA unpacker when duplicated PhotometricInterpretation or BitsPerSample tags are interpreted differently by Pillow and libtiff; the strip path is not affected. No public exploit code or CISA KEV entry was identified at time of analysis, and upgrading to 13.0.0 fixes the issue.
Unrestricted Java deserialization in the MQ Classes for Java shipped with IBM MQ Operator (versions 2.0.0 through 4.0.4) and the long list of MQ Advanced container image builds (9.3.0.x, 9.4.x and 10.0.0.x) lets an authenticated low-privilege user trigger a denial of service against the queue manager and, if a usable gadget chain is present on the classpath, potentially execute arbitrary code. The flaw is rated CVSS 8.8 (AV:N/AC:L/PR:L/UI:N, high confidentiality, integrity and availability impact) and requires valid credentials but no user interaction; the specific interface that accepts the serialized stream is not described in the available sources. There is no evidence of active exploitation - the issue is not in CISA KEV, SSVC records exploitation as none and automatable as no, EPSS is only 0.29% (20th percentile), and no public exploit code has been identified at time of analysis; the vuln.today priority is this_week and IBM has published a vendor fix.
Authenticated users of containerized IBM MQ deployments can cause denial of service or memory corruption by sending MQOPEN operations that are not properly validated for buffer size. The affected deployments identified in the CPE data are IBM MQ Operator and IBM MQ Advanced Container Images, with the provided feed using wildcard version fields rather than a precise version range. The vendor CVSS 3.1 score is 8.8 (AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H), requiring valid MQ credentials and authority to open a queue or object; no public exploit was identified at time of analysis, and a vendor patch is available via IBM support advisory node 7292667.
SQL injection in Integratec's pre-authentication login endpoint (/login/ap_Login.aspx/start) allows remote unauthenticated attackers to read database contents-including credentials and user information-by supplying a crafted UNION-based payload in the sInfo POST parameter. The report carries an 8.8 CVSS score, and our independent assessment is CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:L/A:L, with no authentication or user interaction required; exploitation is limited mainly by whether the login interface is reachable from untrusted networks. No public exploit code was identified at time of analysis, but our independent assessment treats this as a credible, actionable priority rather than a paper-tiger high score.
A stack-based out-of-bounds write in Zephyr's OSDP Control Panel reply parser (cp_decode_response() in subsys/mgmt/osdp/src/osdp_cp.c) lets an attacker on the RS-485 bus overflow fixed 64-byte payload arrays inside a stack-allocated struct osdp_event by declaring an oversized REPLY_KEYPPAD or REPLY_RAW payload length. Any build with CONFIG_OSDP and CONFIG_OSDP_MODE_CP plus an event callback registered is affected, and because nobody validates the reply ID against an outstanding command, a malicious or compromised card reader can answer a routine CMD_POLL with a weaponized frame carrying a correct SOM, valid CRC-16 and matching sequence number. With the default 256-byte UART buffer roughly 180 attacker-chosen bytes land past the event object, corrupting saved registers and the return address on typical Cortex-M/RISC-V layouts, giving crash-to-potential-control-flow-hijack in the OSDP receive thread; no public exploit code has been identified at time of analysis and the issue is bus-adjacent (AV:A) rather than network-reachable.
SQL injection in SLiMS (Senayan Library Management Systems) 9 Bulian 9.8.0 allows an authenticated staff/librarian account to extract arbitrary database contents through the opac_hide and promoted GET parameters of admin/modules/bibliography/index.php, exposing patron records, credentials, and catalogue data. The flaw is reachable by any staff user with access to the bibliography administration module (PR:L), so it is not anonymously exploitable, and the primary impact per our assessment is confidentiality loss rather than data tampering. Publicly available exploit code exists (including a dedicated PoC repository) and CISA SSVC flags exploitation as 'poc', but EPSS remains low at 0.20% (10th percentile) and no active exploitation has been confirmed.
Cross-site request forgery in SLiMS 9 Bulian 9.8.0 allows a remote attacker to trick a logged-in administrator into issuing a GET request to admin/modules/circulation/loan_history_maintenance.php?action=run, which deletes loan history records. The endpoint lacks a CSRF token and POST confirmation, so the forged request executes with the administrator's privileges; the assessed impact is integrity-only (C:N/I:H/A:N), not confidentiality or availability. Exploitation requires an active admin session and user interaction (UI:R); a public PoC reference exists but is not verified as working exploit code, and EPSS is low at 0.20% (9th percentile).
Remote code execution in Craft CMS 5.0.0 through 5.10.x lets authenticated Control Panel users escape the Twig sandbox by supplying attacker-controlled 'as' or 'on' keys to Asset::getMimeType transforms, which are passed unfiltered into Craft::createObject and can attach malicious behaviors or event handlers that execute as the PHP web user. The CVSS 4.0 score is 8.7 (and the assessed CVSS 3.1 vector is AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H), reflecting a high-integrity and high-availability impact once a Control Panel session is obtained. No public exploit code has been identified at the time of analysis and there is no CISA KEV confirmation, so the practical risk is concentrated in environments where Control Panel access is broadly granted, weakly protected, or susceptible to credential phishing or reuse; the fixed release is 5.11.0.
Authenticated remote code execution in Craft CMS 4.x before 4.18.8 and 5.x before 5.11.0 allows control panel users to write a webshell and execute code as the web user by abusing the Twig group filter's handling of non-string arrow arguments as callables. The flaw is a missing checkArrow guard that other arrow-accepting filters apply, letting a crafted element-select condition rule pass an array-callable such as FileHelper::writeToFile. Exploitation requires valid control panel credentials with permission to create or edit condition rules and a writable webroot; no public exploit has been identified at time of analysis.
Authenticated administrator accounts on Craft CMS 5.8.0 through 5.10.x can achieve remote code execution by submitting a crafted JSON-string field layout to ContentBlock::setFieldLayout() that embeds Yii 'as X' or 'on X' configuration keys, which slip past the platform's cleanseConfig() sanitization and let an attacker attach arbitrary behavior classes and event handlers that run OS commands as the web user and expose stored secrets. Exploitation is gated on the allowAdminChanges setting being enabled and on the attacker already holding admin control-panel access, so hardened production installs with that setting disabled (the recommended posture) are not reachable and the realistic attack surface is materially narrower than the 8.7 score suggests; no public exploit has been identified at time of analysis and no CISA KEV entry exists.
Unauthenticated arbitrary code execution in Dell Application Store affects every release prior to 26.05.0.7, reachable by an attacker who shares the same local network segment as the host. The root cause is a critical function that performs no authentication check (CWE-306), so an attacker with adjacent access (shared LAN or Wi-Fi broadcast domain, per AV:A) can invoke it without credentials or user interaction, yielding high confidentiality, integrity, and availability impact (CVSS 8.8). There is no public exploit identified at time of analysis and the issue is not in CISA KEV; the adjacency requirement keeps it off routed/internet attack paths and out of mass-scanning reach, but it is a credible priority for environments running the component on flat or untrusted Layer 2 networks.
Unauthenticated deserialization of attacker-controlled messages in Apache Camel's camel-jgroups consumer can lead to remote code execution in the Camel process. Any host that can reach the JGroups cluster's multicast group or bind port - adjacent network only, given the default UDP multicast stack - can join without credentials and deliver a crafted serialized object that JGroups deserializes on its own receive thread before Camel's routing logic ever runs. Code execution still depends on a usable gadget chain being present on the application classpath and on no JVM-wide -Djdk.serialFilter being configured, which is why the issue is assessed as high rather than critical; no public exploit code or confirmed active exploitation was identified at time of analysis. Deployments that explicitly configure a channel/channelProperties with AUTH and encryption, or that set a JVM-wide serialization filter, are not exploitable through this path.
Expression language injection in the Apache Camel dynamic-router-control endpoint allows a party who can place a message on a transport feeding that endpoint to choose both the expression language and the predicate expression that Camel compiles and then evaluates against every exchange routed through the corresponding dynamic-router endpoint, running with the full privileges of the Camel process - effectively code execution. The issue affects Camel 4.4.0 through 4.18.4 and 4.19.0 through 4.22.0 (releases before 4.4.0 are not affected), scored CVSS 3.1 8.8 (AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H), so some access to the feeding transport is expected but no application-level authentication or user interaction is required. Exposure is limited to deployments that expose the control endpoint through a route forwarding messages from an external, shared or untrusted transport; no public exploit code was identified at time of analysis and the issue is not listed as confirmed actively exploited (CISA KEV).
Deserialization of untrusted data in the Apache Camel Spring Redis component (camel-spring-redis) allows an authenticated attacker who can either PUBLISH to a channel the route subscribes to, or write attacker-controlled bytes into a key the component later reads, to instantiate an arbitrary class graph in the application's JVM and execute its readObject method. Every release from 2.11.0 up to (but not including) 4.18.5 and 4.22.1 ships the affected default, which lazily builds a bare JdkSerializationRedisSerializer with no java.io.ObjectInputFilter whenever no serializer option is configured. No public exploit code was identified at time of analysis, but exploitation needs only a viable gadget on the classpath, which is why this carries CVSS 8.8 with high confidentiality and integrity impact.
PHP object injection in the WordPress plugin Quiz And Survey Master (quiz-master-next) through version 11.2.7 allows an authenticated low-privilege user, such as a subscriber-level account, to pass attacker-controlled data into an unsafe unserialize() call and inject arbitrary PHP objects, with a CVSS 3.1 base score of 8.8 (AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H). Because the vector requires PR:L, the flaw is not reachable by fully anonymous visitors, but any site that allows open user registration effectively lowers the bar to a self-service account. Successful escalation to code execution or data tampering depends on a usable POP gadget chain being present in the plugin, WordPress core, or another installed plugin or theme; no public exploit code or confirmed active exploitation was identified at time of analysis, and the plugin is widely deployed across WordPress sites, making it an attractive target for follow-on research.
Privilege-gated remote code execution in Craft CMS 4.x before 4.18.8 and 5.x before 5.11.0 allows users with template-execution rights-such as Control Panel administrators or any account permitted to author or execute Twig templates-to bypass the Twig create() blocklist by supplying case-varied class names such as 'splfileobject'. Successful abuse can instantiate PHP's SplFileObject to read .env secrets or write a PHP file into the webroot, leading to remote code execution. The official CVSS 4.0 vector lists 8.7 with PR:N, but the independent assessment treats this as privilege-gated (CVSS 3.1 PR:H) and not exploitable by remote unauthenticated attackers against a default install; no public exploit was identified at time of analysis.
Memory corruption in Pillow versions prior to 13.0.0 can crash the process when Image.transform() takes the fast Transform.EXTENT or pure scale-and-translate Transform.AFFINE path with default NEAREST resampling and a selective allocation failure occurs under memory pressure. The defect is a double free/use-after-free in ImagingScaleAffine: the output image allocation succeeds, the additional per-column lookup allocation fails, and the error path deletes image storage that later cleanup releases again (CWE-416). The assessed impact is availability-only with no confidentiality or integrity effect, exploitation depends on attacker-influenced transform sizes plus genuine memory pressure, and no public exploit identified at time of analysis.
Process crash in Pillow before 13.0.0 can be triggered when an application decodes an attacker-supplied PPM or PGM image with zero width and then applies an ImageMorph morphology operation. The flaw is an out-of-bounds memory access in the native morphology routine, and unauthenticated remote attackers can exploit it only when that specific ImageMorph code path is used, resulting in denial of service rather than data exposure or code execution. Vendor-released patch: 13.0.0; no public exploit identified at time of analysis.
Prototype pollution in @rvf/set-get before 7.0.4 lets remote attackers crash a Node.js process by submitting a form field named 'hasOwnProperty.call' (or a similar dotted path into an inherited Object.prototype method) through @rvf/core form parsing. Because setPath resolves inherited properties with a plain `obj[part] === undefined` check, it walks into Object.prototype and overwrites the shared built-in function, denying service process-wide until restart. The CVSS 4.0 vector scores 8.7 with PR:N and UI:N, meaning no authentication or user interaction is required; no public exploit code was identified at time of analysis, and the practical scope is limited to applications that route untrusted form input through the vulnerable @rvf parsing stack, with impact confined to a recoverable denial of service rather than code execution or data exposure.
Account takeover in Craft CMS 5.0.0 through 5.10.x: the users/verify-email action fails to invalidate a user's emailed verification code for active accounts, so a code that has already been used to confirm an email change can be replayed against users/set-password to reset the victim's password. According to the CVSS vector no authentication is required (PR:N), but practical exploitation is gated on obtaining the victim's one-time code - via email interception or control of the change flow - and on the victim actually completing the verify-email step, which is why the assessed vector carries AC:H/UI:R and the issue is considered targeted rather than broadly remote. No public exploit code or confirmed active exploitation was identified at time of analysis, and the behavior is corrected in Craft CMS 5.11.0.
Denial of service in Grav CMS 1.7.53 before 1.7.53.5 and 2.0.0-rc.8 before 2.2.2 lets unauthenticated remote attackers exhaust server memory by bypassing the system.images.max_pixels safeguard when system.images.url_actions is enabled. Attackers use single-dimension or percentage forceResize, zoomCrop, or chained image actions to force PHP's GD library to allocate multi-gigabyte canvases and take the site down. No public exploit identified at time of analysis; the assessed CVSS 3.1 vector (AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H) confirms this is an availability-only threat requiring the url_actions setting to be enabled.
Uninitialized heap memory disclosure in Pillow 5.3.0 through 12.x allows remote attackers to read stale process memory by supplying a crafted TIFF that uses JPEG (COMPRESSION=7) or Group 4 fax compression and triggers a libtiff 'short decode', after which the library converts rows from a buffer whose unwritten tail was never cleared. The tile decode path has been affected since 5.3.0 and the strip path since 6.0.0, and the impact is confidentiality-only (memory disclosure with no corruption, crash, or code execution) - an application is only exposed if it both decodes the attacker-supplied image and returns or re-encodes the resulting pixel data, since that output channel is the only way the stale bytes reach the attacker. No public exploit code or active exploitation is indicated in the available data, and the issue is fixed in Pillow 13.0.0.
Denial of service in Pillow (Python imaging library) versions 5.1.0 through 12.x allows remote attackers to crash or hang applications that append images to attacker-influenced PDFs. PdfParser lacks cycle and duplication tracking when following page-tree /Kids references and only guards against exact repeated trailer offsets, so a crafted PDF with a deep or cyclic page tree, repeated non-cyclic /Kids entries, or a long chain of unique previous-trailer offsets can raise an uncaught RecursionError or drive super-linear CPU and memory consumption before any output is written. Exploitation requires the target application to call Image.save with PDF append enabled against a PDF the attacker can supply; no public exploit code has been identified at time of analysis and the issue is not in CISA KEV, though the availability-only impact and trivial prerequisites justify prompt patching to 13.0.0.
Denial of service in Pillow 5.1.0 through 12.x: when an application uses Image.save with save_all=True to convert a multi-frame image (APNG, GIF, TIFF) into a multipage PDF, PdfParser.next_object_id() rebuilds and rescans the entire set of previously reserved PDF object identifiers for every object, producing quadratic runtime in the number of frames. Because the cost scales with the declared frame count, an unauthenticated remote attacker can submit a very small APNG that advertises a huge number of frames and exhaust CPU on the converting service. Active exploitation has not been confirmed (no CISA KEV listing) and no public exploit code was identified at time of analysis, but the trigger is trivial for anyone who can reach an image-to-PDF conversion workflow; the vendor fixed it in Pillow 13.0.0 via GHSA-jg86-v3jv-hrpm.
Memory exhaustion denial of service in Pillow 4.1.0 through 12.x allows unauthenticated remote attackers to exhaust server memory when an application decodes a crafted MSP image and loads its pixels. The MSP RLE decoder writes expanded run-length output into an in-memory buffer without limiting it to the declared image dimensions, bypassing Pillow's decompression-bomb pixel-count guard; no public exploit identified at time of analysis.
Denial of service in Pillow 5.1.0 through versions before 13.0.0 allows unauthenticated remote attackers to exhaust a processing worker's CPU by supplying a crafted PDF whose page tree contains a Parent-reference cycle when an application uses Pillow to append an image to that PDF. The flaw is an uncontrolled infinite loop in PdfParser.rewrite_pages(), giving availability-only impact with no confidentiality or integrity effect. Vendor-released patch: 13.0.0; no public exploit identified at time of analysis.
Denial of service in Pillow 12.0.0 through 12.x lets unauthenticated remote attackers crash a Python process by supplying a crafted IPTC image that reaches Image.merge(), or by calling Image.merge() with an I/F first band backed by image32 storage. The flaw is a null-pointer dereference (CWE-476) with high availability impact but no confidentiality or integrity impact; it is fixed in 13.0.0 and no public exploit was identified at time of analysis.
Memory exhaustion denial of service in Pillow 9.2.0 through pre-13.0.0 can be triggered by a crafted multi-frame GIF that leaves rejected, oversized frame dimensions stored after Image.DecompressionBombError is raised. The flaw is exploitable only in applications that catch that exception and then continue calling seek() or load() on the same GifImageFile object; applications that let the exception propagate or discard the image object are not affected. It is an availability-only issue, with no public exploit identified at time of analysis and no CISA KEV confirmation.
Memory exhaustion in Pillow 8.1.0 through 12.x allows unauthenticated remote attackers to cause denial of service by supplying a crafted APNG that triggers a full-canvas background allocation inside Image.open() before the decompression-bomb check. The flaw is reachable when an application parses untrusted images with an affected version; no pixel loading, frame iteration, authentication, or user interaction is required. No public exploit code or CISA KEV entry was identified at time of analysis, but the availability impact is high (CVSS 8.7; assessed CVSS:3.1 A:H).
Information disclosure in Grav CMS 2.0.0 through 2.2.1 allows authenticated page editors to expose protected configuration - including security.salt, plugin credentials, and Redis passwords - to every visitor of the affected page by inserting config-*@ directives into a page's frontmatter form. The bypass defeats the Twig sandbox's config_denied_paths redaction, so values the CMS intends to keep internal are rendered into public page output. Our assessment rates this CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:N rather than the advisory's 8.7 CVSS:4.0 vector, because exploitation requires an account already holding page-edit rights; no public exploit code or confirmed active exploitation was identified at time of analysis, and real-world impact depends on whether the deployment actually stores secrets in flat-file configuration.
Authenticated Control Panel users on Craft CMS 5.0.0 through 5.10.x can escape the Twig sandbox and achieve remote code execution by supplying uncleansed criteria to the getNext, getPrev, or getOwners element-query methods. The injected Yii 'as' or 'on' configuration keys attach a malicious behavior or event handler that executes as the web server user when the query runs. No public exploit code or CISA KEV entry was identified at time of analysis; the primary barrier is the requirement for a valid Control Panel account.
Denial of service in picomatch through 4.0.7: a malicious glob pattern such as +(*) or +(*a) compiles into a catastrophically backtracking regular expression that blocks the Node.js event loop and hangs the host application. The flaw sits in analyzeRepeatedExtglob in lib/parse.js, which skipped single-branch repeated extglobs and therefore missed the nested unbounded quantifiers that cause exponential blowup, so risk concentrates in applications that feed attacker-controlled glob patterns into picomatch or its consumers (micromatch, globby, chokidar) with extglob support left at its default. No public exploit code and no CISA KEV entry were identified at time of analysis, and an upstream fix is available as PR #194; fixed-pattern-only usage is not exploitable even though the assessed vector is PR:N/UI:N.
Remote code execution in Semaphore UI through 2.19.16 allows an authenticated user holding the task_runner role to escape the intended Ansible execution boundary by placing protected connection variables into a task's environment JSON. Supplying ansible_ssh_common_args with a ProxyCommand, or ansible_connection=local, makes Ansible run attacker-chosen commands as the Semaphore server process user, exposing encryption keys, SSH keys and other stored secrets. The project addressed the issue upstream in pull request #3888, but no tagged patched release number is stated in the available data, and no public exploit code was identified at time of analysis (CVSS 8.7 / assessed CVSS 3.1 8.8, CWE-915).
An authentication bypass in Eptura Archibus lets remote attackers execute workflow rules without any credentials by abusing the product's DWR (Direct Web Remoting) endpoint: a request is crafted so that the unauthenticated isPasswordUppercase method is invoked first and a workflow-rule execution method is chained as the subsequent call in the same request. The assessed vector (CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:L/A:N, 8.6) reflects a network-reachable, unauthenticated, no-interaction attack with high confidentiality impact and limited integrity impact, plus a scope change because the authentication weakness in the DWR layer lets the attacker reach workflow-rule functionality. No public exploit code and no confirmed active exploitation (CISA KEV) were identified at time of analysis; the primary practical constraint is whether the DWR interface is reachable from the attacker's network position.
OS command injection in television (tv) 0.15.0 through 0.15.9 lets an attacker who can plant a file whose name contains an apostrophe execute arbitrary shell commands with the victim's privileges once that entry is selected and an action is triggered - most notably the bundled default edit action. The vulnerability is local and requires victim interaction (CVSS 4.0 8.5; independently assessed as CVSS 3.1 AV:L/AC:L/PR:N/UI:R with high confidentiality, integrity and availability impact on the victim's account), and no public exploit identified at time of analysis and the issue is not listed in CISA KEV. Upstream fix commits addressing the quoting logic have been published by the vendor.
SQL injection in NocoBase's @nocobase/plugin-collection-sql through 2.2.28 allows an authenticated user with the collection-sql snippet permission to bypass checkSQL() through the sqlCollection:setFields action and persist malicious SQL that executes when the collection reloads. The issue requires high privileges and a PostgreSQL backend for the described impact, where injected statements can read pg_shadow password hashes or invoke pg_read_file; no public exploit was identified at time of analysis. It is a conditional, high-privilege risk rather than an unauthenticated emergency.
Database credential disclosure in Eptura Archibus lets a low-privilege authenticated user retrieve usernames and passwords by submitting a DWR request that names a specific workflow rule; when chained with the companion authentication-bypass flaw, the same primitive enables unauthenticated database reads. The vendor has released a fix referenced by Rapid7, but exploitation requires network reachability to an exposed DWR interface. No public exploit code or confirmed active exploitation has been identified at time of analysis, though the high confidentiality impact and chainability make this a genuine priority for exposed Archibus deployments.
Authenticated stack-based buffer overflow in TP-Link Tapo C320WS V2.0 cameras lets an attacker on the adjacent network who already holds valid device credentials overwrite stack memory through an oversized encrypted password field submitted to the local onboarding Wi-Fi configuration path, ultimately executing commands as root. Because the trigger lives in the provisioning endpoint that is only reachable on the local or onboarding LAN (not from the internet) and requires high-privilege authentication, exploitation is conditionally gated rather than broadly opportunistic; no public exploit code or confirmed active exploitation was identified at time of analysis. Successful abuse lets the attacker take control of the camera, alter its behavior, read sensitive local configuration, and disrupt availability, and TP-Link has published fixed firmware on its download portal.
Local code execution in MobaXterm Personal Edition (Portable) 26.4 GA arises because the application extracts the bundled CygUtils64.plugin without verifying its integrity, allowing an attacker who can write to that file to substitute the contained bin/cygwin1.dll with a malicious library. When the victim subsequently opens a Local Shell session or invokes a Cygwin-based function, the tampered DLL is extracted and loaded in the victim's user context, giving the attacker code execution with the victim's permissions. The flaw requires pre-existing local write access and victim interaction, does not itself elevate privileges, and no public exploit was identified at time of analysis.
BentoML through 1.4.39 renders attacker-controlled docker.env mapping keys into its legacy base.j2 Dockerfile template without neutralizing embedded newline characters, allowing injected Dockerfile instructions - typically RUN commands - to be executed during `bentoml containerize` builds. A victim must be induced to build or containerize an untrusted bentofile.yaml or imported bento archive, so exploitation is local and interaction-dependent (CVSS:3.1/AV:L/AC:L/PR:N/UI:R) rather than remotely triggerable, but successful builds can expose any secrets mounted into the Docker build context and produce poisoned images that downstream users trust and deploy. No public exploit code or confirmed active exploitation was identified at time of analysis; realistic exposure is limited to teams that containerize third-party or otherwise untrusted bento content.
PHP code injection in sqlparse through 0.6.0 can turn crafted SQL text into executable PHP when an application uses the non-default output_format='php' formatter and later runs the generated snippet. The flaw is CWE-94: OutputPHPFilter fails to escape literal dollar signs inside double-quoted PHP strings, so ${expr} or {${expr}} sequences can be interpreted as PHP interpolation. Although the nominal severity is high, exploitation requires a niche formatter plus execution of the generated PHP, and no public exploit was identified at time of analysis.