CVE Feed

Unitree G1 EDU firmware through 1.5.2 contains an unauthenticated remote code execution vulnerability that allows network-adjacent attackers to execute arbitrary commands as root by chaining three weaknesses: an unauthenticated WebRTC-to-DDS bridge on TCP port 9991, a static AES-128 key stored with world-readable permissions, and a path traversal flaw in the chat_go knowledge upload API. Attackers can publish DDS control messages to restart the bashrunner service, plant a malicious payload in its script execution directory via path traversal, and trigger execution of that payload as uid 0 through the bashrunner shell subprocess.

LimeSurvey Community Edition 7.0.5 contains an authenticated improper authorization vulnerability in the survey menu entry creation endpoint. An authenticated user with only the global settings:read permission can directly invoke POST /index.php/admin/menuentries/sa/create and create new survey menu entries without the expected settings:update privilege. The endpoint also allows the attacker to submit menu IDs that the normal interface and intended update workflow restrict for non-superadministrators, enabling unauthorized changes to administrative navigation records. This issue affects LimeSurvey: 7.0.5.

Trilium is an open-source hierarchical note-taking application. In versions up to and including 0.103.0, the default-on "Safe import" filter does not sanitize note titles, and the GeoMap note view interpolates a marker note's title into raw HTML that is rendered as innerHTML, allowing an attacker-supplied import archive to inject script that runs when the map is displayed. Because Safe import neutralizes scripts but never escapes titles, a note whose title contains an HTML event-handler payload survives the import and executes as soon as the victim opens the GeoMap that renders its marker. On the desktop client the Electron renderer runs with Node integration enabled, so the injected JavaScript escalates from cross-site scripting to full remote code execution on the victim's machine. This issue is fixed in version 0.104.0.

Grafana Alloy’s prometheus.operator.servicemonitors component allows a user who can create or modify ServiceMonitor resources in a watched namespace to specify an arbitrary local file through bearerTokenFile. Alloy reads the file and sends its contents as a bearer token to an attacker-controlled scrape endpoint. This may disclose files accessible to the Alloy process, including its projected Kubernetes service account token, potentially granting the attacker Alloy’s Kubernetes permissions. Exploitation requires ServiceMonitor write access and lower privileges than Alloy’s service account.

libreoffice-convert is a Node.js module for converting office documents to different formats. Prior to 1.8.2, index.js uses the caller-controlled options.fileName value in path.join(tempDir.name, fileName) without reducing it to a base name. A filename containing ../ can escape the temporary directory because path.basename() normalization is missing and write the supplied document buffer to an arbitrary path writable by the process, including an SSH authorized_keys file, a cron configuration, or a web root. This issue is fixed in version 1.8.2.

Silverstripe Advanced Workflow is a highly configurable step-based workflow module. Prior to 6.4.5, 7.1.3, and 7.2.1, an attacker with permission to author the advanced workflow email template can place a specially crafted server-side template payload in NotifyUsersWorkflowAction.EmailTemplate. When NotifyUsersWorkflowAction renders the field through the Silverstripe template engine SSTemplateParser, the payload can cause PHP evaluation and arbitrary code execution on the server; the regression coverage is in tests/php/WorkflowEngineTest.php. This issue is fixed in versions 6.4.5, 7.1.3, and 7.2.1.

CC: Tweaked is a mod for Minecraft which adds programmable computers, turtles, and more to the game. Prior to 1.120.0, the SSRF protection in projects/core/src/main/java/dan200/computercraft/core/apis/http/options/AddressPredicate.java blocks the RFC 6052 64:ff9b::/96 NAT64 prefix but omits the RFC 8215 64:ff9b:1::/48 local-use prefix. On a dual-stack server using RFC 8215 NAT64, an unauthenticated user who can execute Lua code can use http.request or http.websocket with an address under 64:ff9b:1::/48 to reach loopback, RFC 1918, cloud metadata, or internal API endpoints because PrivatePattern.matches() does not classify the mapped IPv6 address as private. This issue is fixed in version 1.120.0.

n8n-nodes-sqlite3 is a node for operating a local SQLite database from n8n. Prior to 1.0.0, nodes/SqliteNode/v1/SqliteV1.node.ts exposes the db_path database file path as a node parameter that permits data expressions from upstream workflow input. A workflow author who maps untrusted input to db_path can allow a remote attacker to select which SQLite file the n8n process opens, enabling traversal outside the intended database location and potentially reading, creating, or overwriting files accessible to the process. This issue is fixed in version 1.0.0.

The Tutor LMS WordPress plugin before 4.0.6 does not prevent request data from overwriting internal variables while rendering templates, allowing unauthenticated users to invoke arbitrary zero-argument PHP functions and receive their output.

CakePHP Queue is a queue-interop compatible queueing library. From 0.1.11 until 2.3.1, QueueManager::getUniqueId() generates identifiers for jobs with shouldBeUnique enabled from the job class, method, and parameters, but sorting parameter values drops associative-array keys. An unauthenticated attacker who can influence job parameters can submit semantically different data that produces the same identifier, resulting in legitimate jobs dropped as duplicate collisions. This issue is fixed in version 2.3.1.

EAZ EazyFix 12.9 allows a Security Feature Bypass related to a "Missing Cryptographic Step" associated with "Secure Boot disable."

Silverstripe UserForms provides a visual form builder for the Silverstripe CMS. From 6.0.0 until 6.4.9, 7.0.7, and 7.1.1, the userform email recipient subject field in the CMS accepts a specially crafted payload that can be interpreted as executable server-side code. An authenticated CMS user with permission to configure a UserForms email recipient can use the subject field to run arbitrary code on the server, compromising confidentiality, integrity, and availability. This issue is fixed in versions 6.4.9, 7.0.7, and 7.1.1.

Trilium is an open-source hierarchical note-taking application. In versions prior to 0.104.0, the default-on "Safe import" filter fails to neutralize the shareTemplate relation because that relation is not marked as dangerous, allowing an attacker-supplied import archive to plant a server-side template that leads to remote code execution. The relation is omitted from the built-in list of dangerous attributes, so unlike other code-loading relations it is not disabled on import, and when the victim later publishes the imported note the public share renderer feeds the linked EJS code note's raw bytes into ejs.render, which compiles them in the server's Node process. An unauthenticated request to the shared note then executes the attacker's JavaScript with full access to require, process, the filesystem, and the network. This issue is fixed in version 0.104.0.

Improper Neutralization of Input During Web Page Generation ('Cross-site Scripting') vulnerability in woylie doggo allows Reflected XSS. Doggo.normalize_value/2 in lib/doggo.ex returned date field values wrapped in {:safe, ...}, the Phoenix.HTML marker meaning "already escaped, emit verbatim", without escaping them, so the value reached the value attribute of the <input> rendered by the field component unchanged. Any application rendering <.field type="date"> over user-controlled params is affected through the ordinary Phoenix form round-trip, where a failed validation re-renders the submitted value. The pattern kept exactly the first ten bytes and discarded shorter values, capping a payload at ten bytes: enough to terminate the attribute and open an element or attach a short event handler, not enough to place attacker-chosen script inline. Only type="date" is affected. This issue affects doggo: from 0.1.0 before 0.14.8.

Applications using Spring Framework's FreeMarker integration may be vulnerable to a path traversal attack when a controller returns a view name derived from untrusted input and FreeMarker is configured to resolve templates through SpringTemplateLoader. Spring Framework 7.0.0 - 7.0.8 Spring Framework 6.2.0 - 6.2.19 Spring Framework 6.1.0 - 6.1.28 Spring Framework 6.0.0 - 6.0.30 Spring Framework 5.3.0 - 5.3.49 Spring Framework 5.2.25.RELEASE and earlier

Any application shipping logs to RabbitMQ over TLS via the Log4j2 appender, relying on the documented default, is exposed to man-in-the-middle interception of every log event. Spring AMQP 4.1.0 Spring AMQP 4.0.0 - 4.0.4 Spring AMQP 3.2.0 - 3.2.12 Spring AMQP 2.4.18 and earlier

Reflected XSS in Netron versions <=9.1.2 on desktop application through unsanitized node names allows an attacker to hide certain nodes, perform port scanning or abuse a Chrome n-day to achieve Remote Code Execution.

Reflected XSS in Netron versions <=9.1.2 on desktop application through unsanitized node names allows an attacker to hide certain nodes, perform port scanning or abuse a Chrome n-day to achieve Remote Code Execution.

Reflected XSS in Netron versions <=9.1.2 on desktop application through unsanitized node names allows an attacker to hide certain nodes, perform port scanning or abuse a Chrome n-day to achieve Remote Code Execution.

When the ClickHouse plugin uses Native protocol (the default) with PDC or secure SOCKS, it asks for TLS but the connection library ignores that and talks to ClickHouse in the clear. Username, password, queries, and results can be read on the hop after the proxy. The server certificate is never checked, and a configured client certificate is not sent.

GitLab has remediated a vulnerability in the GitLab AI Gateway component affecting all versions of the AI Gateway from 18.10 to 19.0.12, 19.1 to 19.1.7, and 19.2 to 19.2.2 that could have allowed an authenticated user with Duo Agent Platform access to redirect outbound model requests to an externally-controlled endpoint via a crafted inline flow configuration that overrides the HTTP Host header, resulting in disclosure of Google Cloud Vertex cloud service credentials and private signing keys.

GitLab has remediated a vulnerability in the GitLab AI Gateway component affecting all versions of the AI Gateway from 18.9.0 to 19.0.12, 19.1 to 19.1.7, and 19.2 to 19.2.2 that could have allowed an authenticated user with Duo Agent Platform access to redirect model requests to an externally-controlled endpoint via crafted model metadata, resulting in the disclosure of Google Vertex AI or AWS Bedrock cloud service credentials.

A flaw was found in Undertow. A remote attacker could exploit this vulnerability by sending specially crafted WebSocket messages with permessage-deflate negotiated. This could lead to excessive memory consumption due to the PerMessageDeflateFunction.largerBuffer() method using exponential doubling, resulting in a Denial of Service (DoS) for the affected application.

In MongoDB Connector for BI, mongodrdl may write a TLS private-key password to standard error when the password is supplied through both the connection URI and the corresponding command-line option. A local user with access to the captured command output and encrypted key file may use the disclosed password to access the associated TLS client key.

An unauthenticated client that can reach a MongoDB Connector for BI deployment configured with Kerberos authentication may cause mongosqld to terminate when a crafted authentication exchange encounters a specific GSSAPI error-handling condition. This can interrupt BI Connector availability until the process restarts.

Substance3D - Sampler versions 5.1.3 and earlier are affected by a Heap-based Buffer Overflow vulnerability that could result in arbitrary code execution in the context of the current user. Exploitation of this issue requires user interaction in that a victim must open a malicious file.

The Twig sandbox mechanism in Craft CMS is configured to allow dangerous functionality from the Yii framework, leading to authenticated RCE similar to previously disclosed vulnerabilities.

An out-of-bounds read was found in the DHCPv4 packet capture code of wicked. ni_capture_inspect_udp_header() in src/capture.c reports the IP total length as the payload length instead of the length of the remaining UDP payload. Consequently, the DHCP option walker in the DHCPv4 client (wickedd-dhcp4) reads up to ihl + 8 bytes — at most 68 bytes — past the end of the 1500-byte packet receive buffer. An unauthenticated attacker on the same network who sends a crafted DHCP/UDP packet can make the client parse adjacent heap memory as DHCP options, so that heap contents such as allocator metadata or pointer values can be interpreted into lease fields. The over-read is bounded to 68 bytes; no memory write, no attacker control over the adjacent bytes and no remote exfiltration primitive has been demonstrated. This issue affects wicked up to and including version 0.6.80.

An integer underflow was found in the DHCPv4 packet capture code of wicked. ni_capture_inspect_udp_header() in src/capture.c does not verify that the IP total length field (ip_len) is at least as large as the IP header length (ihl) before subtracting the header length. An unauthenticated attacker on the same network can thereby trigger an out-of-bounds read past the receive buffer in the wicked DHCPv4 client (wickedd-dhcp4), which can crash the daemon depending on the process memory layout. No information disclosure has been demonstrated. This issue affects wicked up to and including version 0.6.80.

startServer.ts in the mcp-http-server package of UI-TARS-desktop defaulted its listen address to '::' when no host was given, so startSseAndStreamableHttpMcpServer bound the Streamable HTTP and SSE MCP transports to every interface, and its authentication middleware was optional: middlewares are applied only when a caller supplies them. The @agent-infra/mcp-server-commands and @agent-infra/mcp-server-filesystem entry points call startSseAndStreamableHttpMcpServer with a host and port alone and pass no middleware, so neither server required a credential. The commands server exposes a run_command tool that hands its caller-supplied command string to promisify(child_process.exec), so any unauthenticated client able to reach the port could run arbitrary commands as the user running the server, and the filesystem server exposed its file read and write tools on the same terms. The listen default became 127.0.0.1 in commit c2ad42e3eb9b27830db41a3e6f51ca7179d9b168; the package version stayed at 1.2.4 across that change, so the boundary is the commit rather than a release.

NLTK versions before 3.10.3 contain a filesystem containment bypass vulnerability in the Downloader.download and Downloader.incr_download methods that allows attackers to overwrite files outside the install root through pre-existing hardlinks. Attackers with write access to a shared downloader directory can create hardlinks pointing to outside-root files that are then overwritten during normal package extraction, mutating files outside the intended install tree.

NLTK through 3.10.3 contains a path traversal vulnerability in model-artifact APIs that bypass pathsec enforcement by using raw file operations on caller-controlled paths. Attackers can read or write files outside allowed sandbox roots through TransitionParser, AveragedPerceptron, PerceptronTagger, and maxent parameter APIs when pathsec is enabled.

NLTK before 3.10.3 contains a regular expression denial of service vulnerability in Pl196xCorpusReader that allows attackers to cause quadratic CPU consumption by supplying malformed TEI blocks with many unmatched opening tags. Attackers can exploit lazy regex patterns in the read_block method through public APIs like words() and tagged_words() to force repeated rescans and achieve near-quadratic runtime growth.

NLTK before 3.10.3 contains an uncontrolled recursion vulnerability in nltk.featstruct.FeatStructReader that allows unauthenticated attackers to cause a denial of service by supplying deeply nested feature-structure input. Attackers can craft trivial payloads with nested brackets that exceed Python's recursion limit and raise an unhandled RecursionError, crashing applications that parse user-supplied feature structures or feature grammars.

NLTK versions before 3.10.3 contain a quadratic CPU exhaustion vulnerability in XMLCorpusView._read_xml_fragment() that rescans accumulated XML fragments on every 1 KiB block read. Attackers can provide malformed XML corpus files to cause severe CPU consumption and denial of service through affected readers like BNCCorpusReader.

openssl_encrypt before 1.4.9 fails to validate KDF cost parameters in encrypted file metadata and keystore headers, allowing attackers to trigger unbounded memory allocation. Attackers can craft malicious encrypted files declaring arbitrarily large Argon2, scrypt, or balloon KDF parameters to exhaust system memory and crash the process without authentication.

nltk PorterStemmer in versions <= 3.10.2 (fixed in 3.10.3) contains an inefficient-algorithmic-complexity denial of service in PorterStemmer.stem(). The _is_consonant() helper walks backward over the entire run of trailing 'y' characters on every call, and _measure() invokes it for each stem position, causing O(n^2) behavior. A single ~20-50 KB untrusted token consisting of a long run of the letter 'y' followed by a matching suffix (e.g., 'ness') can pin a CPU core for seconds to minutes, causing availability impact.

openssl_encrypt before 1.4.9 fails to validate the memory_cost parameter from identity file protection blocks, allowing attackers to trigger out-of-memory conditions during key derivation. Attackers with write access to local identity stores can craft malicious identity files with excessive memory_cost values that cause the host to crash when unlocking identities before authentication.

openssl_encrypt before 1.4.9 executes untrusted third-party plugins with insufficient controls: the plugin signature policy defaulted to WARN, so an unsigned/unverifiable non-built-in plugin was compiled and executed in the host process at import time, before the runtime sandbox is installed. The only default gate was an incomplete, bypassable AST denylist. If a user is induced to load an attacker's plugin, this results in arbitrary code execution with the privileges of the user running openssl_encrypt. Fixed in 1.4.9 by defaulting the signature policy to ENFORCE for non-built-in plugins.

openssl_encrypt versions before 1.4.9 use under-parameterized PBKDF2-HMAC-SHA256 with only 100,000 iterations to protect PQC keyfile private keys and 10,000 iterations for dual-encryption file-password verification. Attackers who obtain keyfiles or encrypted files can brute-force wrapping passwords offline using GPU or ASIC acceleration.

openssl_encrypt (pip package openssl-encrypt) before 1.4.9 contains two weaknesses in the portable USB drive feature, whose threat model treats the removable drive as untrusted (attacker with physical write access). USBDriveCreator._verify_integrity_file only validates files listed in the manifest, so files added to the drive — including a root-level autorun payload — are not detected and integrity verification still passes. Additionally, a globally constant, source-embedded KDF salt (_LEGACY_FIXED_SALT) is used to derive the drive encryption key for any drive lacking a per-drive salt file, defeating precomputation resistance and enabling an offline rainbow-table attack.

openssl_encrypt (pip: openssl-encrypt) versions before 1.4.9 contain a path traversal flaw in PluginSandbox._is_safe_path, which authorized file access using a bare string-prefix match. A sandboxed plugin without the READ_FILES permission could read or write another plugin's directory that merely shares a name prefix (e.g., .../plugins/foobar matching allowed .../plugins/foo), breaking per-plugin isolation within the same user. Fixed by matching each allowed directory exactly or with a trailing path separator.

openssl_encrypt (pip package openssl-encrypt) versions <= 1.4.8 do not redact the keyserver bearer token passed as the positional argument to 'keyserver set-token' in the --debug argv dump, because sanitize_argv_for_debug fails to sanitize it. As a result the token is printed in cleartext to stderr under --debug (even without --unsafe-show-secrets), persisting the credential in logs and terminal history. Fixed in 1.4.9.

openssl_encrypt (pip: openssl-encrypt) versions <= 1.4.8 use suffix-tolerant fingerprint matching in enroll_trust_key when binding a plugin-signing trust anchor. An operator who confirms a short (forgeable, ~32-bit) GPG key id could unknowingly enroll an attacker's colliding key as a trusted anchor, which then vouches for malicious plugins under the ENFORCE signature policy. Version 1.4.9 fixes this by requiring the confirmed value to exactly match the full primary-key fingerprint (case-insensitive, whitespace-stripped).

openssl_encrypt before 1.4.9 fails to sanitize the email field of imported identity documents, allowing attackers to inject ANSI escape sequences that forge the fingerprint verification line displayed to users. Attackers can deliver a crafted identity bundle through normal contact-exchange flows or keyserver responses to manipulate terminal output and display a fraudulent fingerprint, bypassing the out-of-band verification mechanism that protects against key substitution attacks.

openssl_encrypt before 1.4.9 fails to prevent namespace collisions between own identities and contacts in IdentityStore, allowing attackers to create shadowed contact entries invisible until the corresponding own identity is deleted. When the own identity is deleted, the shadowed contact becomes visible and resolves to the attacker's keys, enabling silent key substitution for encrypted files.

openssl-encrypt before 1.4.9 fails to redact the file password in its --debug argv dump when the password is supplied via bundled short-option spellings (e.g. -apHunter2) or abbreviated long-option spellings (e.g. --passw). The sanitizer only recognized exact option names, --option=value forms, and tokens starting with -p, so these spellings bypass the redaction chokepoint and the cleartext password is written to stderr. Anyone with access to that output (terminal scrollback, merged 2>&1 output, CI job logs, or the GUI's persistent debug log) can recover the password.

openssl_encrypt versions before 1.4.9 contain a weak key derivation vulnerability in the D-Bus CryptoService.EncryptFile handler that uses unstretched SHA-256 instead of Argon2id. Attackers can perform offline password guessing against encrypted files roughly six to seven orders of magnitude faster than documented protection by exploiting the missing key stretching and hash rounds.

openssl_encrypt versions before 1.4.9 fail to validate encryption status of embedded post-quantum private keys in file metadata. Attackers can craft files with unencrypted embedded PQC keys that decrypt under any password, bypassing authentication and producing attacker-chosen plaintext with false integrity verification.

openssl_encrypt before 1.4.9 fails to re-derive and validate fingerprints when loading identities from identity.json, allowing attackers to substitute public keys in identity stores. Attackers can replace legitimate public keys with their own while maintaining the claimed fingerprint, enabling silent key substitution where encryption uses attacker keys and signature verification appears valid.