Vulnerabilities
All
My stack
Known exploited
Public exploit
Critical (9+)
Likely exploited (EPSS ≥ 50%)
Changed, 24 h
New, 24 h
Sort:
risk
modified
published
CVSS
EPSS
| CVE | CVSS | EPSS | Vendor, product | Description | Modified |
|---|---|---|---|---|---|
| CVE-2026-9198KEVexploit ×11 | 9.8 critical | 29.0% | langflow | IBM Langflow OSS 1.0.0 through 1.10.0 allows unauthenticated attackers to chain /api/v1/auto_login (mints SUPERUSER tokens to any network caller) with /api/v1/validate/code (executes user code via exec()) to achieve full RCE on default Langflow deployments | Sat 24 Oct |
| CVE-2025-59536exploit ×6 | 8.7 high | 25.0% | anthropic, claude_code | Claude Code is an agentic coding tool. Versions before 1.0.111 were vulnerable to Code Injection due to a bug in the startup trust dialog implementation. Claude Code could be tricked to execute code contained in a project before the user accepted the startup trust dialog. Exploiting this requires a user to start Claude Code in an untrusted directory. Users on standard Claude Code auto-update will have received this fix automatically. Users performing manual updates are advised to update to the latest version. This issue is fixed in version 1.0.111. | 12 Oct |
| CVE-2025-13390exploit ×2 | 9.8 critical | 5.0% | wpdirectorykit, wp_directory_kit | The WP Directory Kit plugin for WordPress is vulnerable to authentication bypass in all versions up to, and including, 1.4.4 due to incorrect implementation of the authentication algorithm in the "wdk_generate_auto_login_link" function. This is due to the feature using a cryptographically weak token generation mechanism. This makes it possible for unauthenticated attackers to gain administrative access and achieve full site takeover via the auto-login endpoint with a predictable token. | 18 Oct |
| CVE-2026-64824exploit | 9.3 critical | 1.0% | no CPE yet | Home Assistant Core before 2026.7.0 contains a path traversal vulnerability in the backup-restore function that allows attackers to write files to arbitrary absolute filesystem paths by supplying a crafted tar archive with a SYMTYPE entry containing a benign member name paired with an absolute linkname pointing outside the extraction directory. Because the official Docker image runs the Home Assistant process as root and the subsequent regular-file entry is written through the unvalidated symlink, attackers can achieve remote code execution by overwriting auto-imported Python paths such as site-packages/sitecustomize.py or custom component directories. | Sat 24 Oct |
| CVE-2026-47117exploit | 9.3 critical | 1.0% | no CPE yet | OpenMed before 1.5.2 contains a remote code execution vulnerability in the PII privacy-filter model loading path. The privacy-filter dispatcher used broad substring matching on the user-supplied model_name parameter, allowing a value such as attacker/foo-privacy-filter-bar to route through a path that loads Hugging Face models with trust_remote_code=True. An unauthenticated attacker can supply a malicious model repository containing custom Transformers code via auto_map in config.json or tokenizer_config.json, which is imported and executed with the privileges of the OpenMed service process. | 8 Oct |
| CVE-2026-69263exploit | 8.7 high | 1.0% | flowiseai, flowise | Flowise is a drag & drop user interface to build a customized large language model flow. Prior to 3.1.3, the mitigation for CVE-2025-8943 blocked -y and --yes flags on npx, but packages/components/nodes/tools/MCP/core.ts denied only PATH, LD_LIBRARY_PATH, DYLD_LIBRARY_PATH, and NODE_OPTIONS by exact environment-variable name. Because npm reads configuration from npm_config_* variables, setting npm_config_yes=true reproduced --yes behavior without using a blocked flag, causing npx to auto-install and execute the named package when a Custom MCP server launched. This issue is fixed in version 3.1.3. | Sat 24 Oct |
| CVE-2026-54316exploit | 6.0 medium | 1.0% | anthropic, claude_code | Claude Code is an agentic coding tool. From 0.2.54 until 2.1.163, because the hostname huggingface.co was pre-approved as a bare hostname for the WebFetch tool, any path on that domain—including attacker-controlled model repositories—was auto-approved without a permission prompt or being subject to --allowedTools restrictions. An attacker able to inject untrusted content into a Claude Code context could direct it to issue WebFetch requests against attacker-controlled repository files (e.g. /resolve/main/config.json), which HuggingFace counts as downloads server-side, creating a covert out-of-band channel for encoding and exfiltrating data Claude can access such as files, environment variables, or command output. Reliably exploiting this required the ability to add untrusted content into a Claude Code context window. This vulnerability is fixed in 2.1.163. | Sun 25 Oct |
| CVE-2026-97657exploit | 9.8 critical | 0.00% | no CPE yet | The Abandoned Cart Pro for WooCommerce plugin for WordPress is vulnerable to Authentication Bypass via Malleable Encryption and Bypassable HMAC in all versions up to, and including, 10.8.0. This is due to the use of MAC-free AES-CTR encryption for cart-recovery link tokens — where the 8-byte CTR nonce is carried inside the ciphertext blob and the encryption key is looked up using the attacker-supplied `user_email` GET parameter — combined with an HMAC nonce check in `wcap_email_track_links()` that is only enforced when the `wcap_nonce` parameter is non-empty, allowing complete bypass by omission, and an insecure direct object reference on the decrypted `abandoned_id` with no ownership verification confirming the cart belongs to the email used for decryption. This makes it possible for unauthenticated attackers who possess a single legitimate cart-recovery link (obtained from their own reminder email) to recover the keystream, forge a ciphertext that decrypts to any victim's cart ID, and trigger a call to `wp_set_auth_cookie()` for the victim user — achieving full account takeover of any user who has ever abandoned a cart, including administrators. The auto-login feature defaults to enabled and is forced on during plugin upgrades, meaning all affected installations are exploitable without any additional configuration. | 14 Oct |
| CVE-2026-35570exploit | 8.4 high | 0.00% | gitlawb, openclaude | OpenClaude is an open-source coding-agent command line interface for cloud and local model providers. Versions prior to 0.5.1 have a logic flaw in `bashToolHasPermission()` inside `src/tools/BashTool/bashPermissions.ts`. When the sandbox auto-allow feature is active and no explicit deny rule is configured, the function returns an `allow` result immediately — before the path constraint filter (`checkPathConstraints`) is ever evaluated. This allows commands containing path traversal sequences (e.g., `../../../../../etc/passwd`) to bypass directory restrictions entirely. Version 0.5.1 contains a patch for the issue. | Mon 26 Oct |
| CVE-2026-15723exploit | 8.2 high | 0.00% | gitlab | GitLab has remediated an issue in GitLab CE/EE affecting all versions from 10.1 before 19.3.6, 19.4 before 19.4.3, and 19.5 before 19.5.1 that, when OmniAuth SAML and LDAP auto-linking were both enabled, an authenticated user could have taken over an existing LDAP-backed account without any interaction from the targeted user, due to improper input sanitization in the LDAP email lookup path used during OmniAuth identity linking. | Tue 27 Oct |
| CVE-2026-14440exploit | 7.6 high | 0.00% | no CPE yet | Description: To issue and renew TLS certificates on behalf of customers, Cloudflare's Universal SSL feature automatically manages the CAA RRset for the customer's zone. This auto-managed RRset is permissive by design (e.g. 'issue "letsencrypt.org"' without parameters). On Universal SSL zones, Cloudflare's authoritative DNS serves this auto-managed RRset at query time, superseding any customer-configured CAA records on the zone. When a customer publishes a stricter CAA record using the RFC 8657 accounturi or validationmethods parameters, the Certificate Authority does not observe those parameters when evaluating the served RRset under RFC 8659. As a result, the RFC 8657 account-binding and validation-method-binding protections are not enforced end-to-end on Universal SSL zones. Successful exploitation could result in issuance of a browser-trusted TLS certificate to an attacker, enabling MITM against the affected domain. Exploitation is non-trivial in practice: an attacker would need to hold an ACME account at one of the Certificate Authorities in the served CAA RRset and to simultaneously satisfy domain control validation across the multiple geographically distinct Network Perspectives the CA relies on for Multi-Perspective Issuance Corroboration. Cloudflare prefixes are anycast-announced from hundreds of locations globally, raising the bar against single-vantage-point BGP hijacks. Any resulting misissuance of a browser-trusted certificate is subject to Certificate Transparency logging required by major browsers, and would be visible to CT monitoring. Mitigation: Customers requiring strict RFC 8657 enforcement need to disable Universal SSL on the affected zone. Universal SSL's automatic CAA management and customer-set RFC 8657 accounturi and validationmethods enforcement are mutually exclusive by the nature of the issue, so there is no in-product workaround that preserves both. Certificate Transparency monitoring is recommended for all customers as a general detection control. Credits: David Osipov (ORCID: https://orcid.org/0009-0005-2713-9242), independent researcher | Sun 25 Oct |
| CVE-2026-7669exploit | 6.3 medium | 0.00% | no CPE yet | A vulnerability was detected in sgl-project SGLang up to 0.5.9. Impacted is the function get_tokenizer of the file python/sglang/srt/utils/hf_transformers_utils.py of the component HuggingFace Transformer Handler. The manipulation of the argument trust_remote_code with the input False as part of Boolean results in code injection. The attack can be executed remotely. A high complexity level is associated with this attack. The exploitability is considered difficult. In get_tokenizer(), when the caller passes trust_remote_code=False and HuggingFace transformers v5 returns a TokenizersBackend instance (the generic fallback for tokenizer classes not in the registry), SGLang silently re-invokes AutoTokenizer.from_pretrained with trust_remote_code=True, overriding the caller's explicit security setting. A model repository containing a malicious tokenizer.py referenced via auto_map in tokenizer_config.json will execute arbitrary Python in the SGLang process during this second call. No log line or warning is emitted. The override affects all current SGLang versions because transformers==5.3.0 is pinned in pyproject.toml. Both tokenizer_mode="auto" and tokenizer_mode="slow" are affected. The exploit is now public and may be used. The vendor was contacted early about this disclosure but did not respond in any way. | Mon 26 Oct |
| CVE-2026-23007exploit | 5.5 medium | 0.00% | linux, linux_kernel | In the Linux kernel, the following vulnerability has been resolved: block: zero non-PI portion of auto integrity buffer The auto-generated integrity buffer for writes needs to be fully initialized before being passed to the underlying block device, otherwise the uninitialized memory can be read back by userspace or anyone with physical access to the storage device. If protection information is generated, that portion of the integrity buffer is already initialized. The integrity data is also zeroed if PI generation is disabled via sysfs or the PI tuple size is 0. However, this misses the case where PI is generated and the PI tuple size is nonzero, but the metadata size is larger than the PI tuple. In this case, the remainder ("opaque") of the metadata is left uninitialized. Generalize the BLK_INTEGRITY_CSUM_NONE check to cover any case when the metadata is larger than just the PI tuple. | 21 Oct |
| CVE-2026-104587exploit | 5.4 medium | 0.00% | no CPE yet | blob42 gosuki through commit 6e2bf34 contains a Stored Cross-Site Scripting (XSS) vulnerability. The application uses text/template instead of html/template, failing to perform context-aware auto-escaping on the URL and DisplayURL fields of bookmarks in internal/webui/templates/bookmarks.html. An attacker can craft a malicious bookmark URL containing HTML attribute-breaking payloads that execute arbitrary JavaScript in the victim's browser context when rendered in the local web interface. | 16 Oct |
| CVE-2025-65093 | 5.5 medium | 4.0% | librenms | LibreNMS is an auto-discovering PHP/MySQL/SNMP based network monitoring tool. Prior to version 25.11.0, a boolean-based blind SQL injection vulnerability was identified in the LibreNMS application at the /ajax_output.php endpoint. The hostname parameter is interpolated directly into an SQL query without proper sanitization or parameter binding, allowing an attacker to manipulate the query logic and infer data from the database through conditional responses. This issue has been patched in version 25.11.0. | 19 Oct |
| CVE-2025-68614 | 5.4 medium | 4.0% | librenms | LibreNMS is an auto-discovering PHP/MySQL/SNMP based network monitoring tool. Prior to version 25.12.0, the Alert Rule API is vulnerable to stored cross-site scripting. Alert rules can be created or updated via LibreNMS API. The alert rule name is not properly sanitized, and can be used to inject HTML code. This issue has been patched in version 25.12.0. | 19 Oct |
| CVE-2026-9103 | 9.8 critical | 3.0% | langflow | IBM Langflow OSS 1.0.0 through 1.10.0 could allow a remote attacker to gain unauthorized access due to improper authentication in the /api/v1/login/auto_login endpoint. The endpoint issues long-lived superuser bearer tokens without requiring authentication when the AUTO_LOGIN configuration is enabled (enabled by default), which may allow an unauthenticated network attacker to obtain full administrative access. Additionally, permissive cross-origin resource sharing (CORS) settings may allow tokens to be exposed to unintended origins, increasing the risk of unauthorized access. | Sat 24 Oct |
| CVE-2025-34334 | 8.7 high | 3.0% | audiocodes, fax_server, interactive_voice_response | AudioCodes Fax Server and Auto-Attendant IVR appliances versions up to and including 2.6.23 are vulnerable to an authenticated command injection in the fax test functionality implemented by AudioCodes_files/TestFax.php. When a fax "send" test is requested, the application builds a faxsender command line using attacker-supplied parameters and passes it to GlobalUtils::RunBatchFile without proper validation or shell-argument sanitization. The resulting batch file is written into a temporary run directory and then executed via a backend service that runs as NT AUTHORITY\\SYSTEM. An authenticated attacker with access to the fax test interface can craft parameter values that inject additional shell commands into the generated batch file, leading to arbitrary command execution with SYSTEM privileges. In addition, because the generated batch files reside in a location with overly permissive file system permissions, a local low-privilege user on the server can modify pending batch files to achieve the same elevation. | 18 Oct |
| CVE-2025-34335 | 8.7 high | 3.0% | audiocodes, fax_server, interactive_voice_response | AudioCodes Fax Server and Auto-Attendant IVR appliances versions up to and including 2.6.23 expose an authenticated command injection vulnerability in the license activation workflow handled by AudioCodes_files/ActivateLicense.php. When a license file is uploaded, the application derives a new filename by combining a generated base name with the attacker-controlled extension portion of the original upload name, then constructs a command line for fax_server_lic_cmdline.exe that includes this path. The extension value is incorporated into the command string without input validation, escaping, or proper argument quotation before being passed to exec(). An authenticated user with access to the license upload interface can supply a specially crafted filename whose extension injects additional shell metacharacters, causing arbitrary commands to be executed as NT AUTHORITY\\SYSTEM. | 18 Oct |
| CVE-2013-10073 | 8.7 high | 3.0% | nagios, nagios_xi | Nagios XI versions prior to 2012R1.6 contain a shell command injection vulnerability in the Auto-Discovery tool. User-controlled input is passed to a shell without adequate sanitation or argument quoting, allowing an authenticated user with access to discovery functionality to execute arbitrary commands with the privileges of the application service. | 18 Oct |
| CVE-2026-82854 | 9.3 critical | 2.0% | no CPE yet | Nodemailer before 8.0.4 is vulnerable to SMTP command injection through the unsanitized envelope.size parameter. When an application passes a custom envelope object with a size property containing CRLF characters to sendMail(), the value is concatenated into the SMTP MAIL FROM command (as SIZE=...) without sanitization, allowing injection of arbitrary SMTP commands such as RCPT TO to silently add attacker-controlled recipients. Exploitation requires the application to expose the envelope size to attacker-controlled input, as Nodemailer does not include size in the default auto-constructed envelope. | 8 Oct |
| CVE-2026-40502 | 8.7 high | 2.0% | hkuds, openharness | OpenHarness prior to commit dd1d235 contains a command injection vulnerability that allows remote gateway users with chat access to invoke sensitive administrative commands by exploiting insufficient distinction between local-only and remote-safe commands in the gateway handler. Attackers can execute administrative commands such as /permissions full_auto through remote chat sessions to change permission modes of a running OpenHarness instance without operator authorization. | Mon 26 Oct |
| CVE-2026-63108 | 7.7 high | 2.0% | no CPE yet | Roo Code through 3.54.0 contains a command injection vulnerability in the auto-approve execute feature that allows attackers to bypass allowlist/denylist enforcement by nesting command substitutions inside parameter expansion defaults. The command parser in parse-command.ts replaces parameter expansions with opaque placeholders before extracting command substitutions, causing the containsDangerousSubstitution guard to miss nested payloads, which are then auto-approved based on the outer allowlisted command prefix and executed by the shell via execa, enabling arbitrary command execution. | Sat 24 Oct |
| CVE-2026-9135 | 9.9 critical | 1.0% | langflow | IBM Langflow OSS 1.0.0 through 1.10.0 Langflow versions up to 1.9.2 (commit 94981c443d4918517b9e8163d70fc598dc33a32d) contain a code injection vulnerability in the Policies component's ToolGuard integration that bypasses the allow_custom_components=false security control. The vulnerability exists because the validation mechanism only checks the main component source code in node_template["code"]["value"] but fails to validate dynamic CodeInput fields that store generated ToolGuard Python files. Attackers can embed malicious Python code in these unvalidated dynamic fields, which are persisted in Flow.data and later executed server-side when a guarded tool is invoked through the ToolGuard runtime. This allows authenticated users with flow creation privileges to achieve arbitrary Python code execution on the backend despite custom component restrictions. The vulnerability can be escalated through cross-tenant flow manipulation via the agentic MCP update_flow_component_field tool, which accepts attacker-controlled user_id parameters, enabling attackers to inject malicious code into victim users' flows. When combined with publicly accessible flows and specific misconfigurations (AUTO_LOGIN=true, NEW_USER_IS_ACTIVE=true), the attack can be conducted with reduced authentication requirements. | Sat 24 Oct |
| CVE-2026-105697 | 9.9 critical | 1.0% | no CPE yet | Langflow is a tool for building and deploying AI-powered agents and workflows. Before Langflow 1.10.3, the MCP stdio transport launched whatever command / args a user put in an MCP server configuration, with no allowlist and (before 1.10.3) wrapped in bash -c "exec {command} ...". Any user able to reach the MCP server settings ("Settings → MCP Servers → Add MCP Server", POST/PATCH /api/v2/mcp/servers/{server_name}) or to build a flow with the MCP Tools component could add a "server" whose command is an arbitrary OS command (touch, rm -rf, a reverse shell, ...). The command runs on the Langflow host as the Langflow process user as soon as Langflow tries to connect to the server (listing servers, loading tools, running the flow) — even when the UI then reports that the stdio server failed to start. With the default LANGFLOW_AUTO_LOGIN=true, GET /api/v1/auto_login hands out a token without credentials, so on an exposed instance running the default configuration this is reachable without an account. AUTO_LOGIN is documented as a development-only setting; with it disabled, any authenticated (non-admin) user can exploit it. This issue is fixed in Langflow 1.10.3, langflow-base 0.10.3, and lfx 1.10.3. | 9 Oct |
| CVE-2026-9202 | 9.8 critical | 1.0% | langflow | IBM Langflow OSS 1.0.0 through 1.10.0 allows unauthenticated attackers to create unlimited user accounts on any Langflow instance; when NEW_USER_IS_ACTIVE=true (documented deployment option), newly created accounts are immediately active and can authenticate to reach RCE endpoints, bypassing the need for AUTO_LOGIN. | Sat 24 Oct |
| CVE-2026-72191 | 9.8 critical | 1.0% | no CPE yet | In the Linux kernel, the following vulnerability has been resolved: ntfs3: validate split-point offset in indx_insert_into_buffer indx_insert_into_buffer() computes used = used1 - to_copy - sp_size; memmove(de_t, Add2Ptr(sp, sp_size), used - le32_to_cpu(hdr1->de_off)); where sp and sp_size come from hdr_find_split(). hdr_find_split() walks entries by le16_to_cpu(e->size) without validating that each step stays within hdr->used or that the size field is at least sizeof(struct NTFS_DE). index_hdr_check(), the on-load gatekeeper, only validates header-level fields (used, total, de_off) and does not walk per-entry sizes. A crafted NTFS image whose leaf INDEX_HDR reports used == total but contains one interior NTFS_DE with size = 0xFFF0 therefore passes validation, descends to indx_insert_into_buffer() through the ntfs_create() -> indx_insert_entry() path, and makes hdr_find_split() return an sp whose sp_size (0xFFF0) greatly exceeds the remaining bytes in the buffer. The u32 subtraction underflows and the memmove count becomes a near-4-GiB value, producing an out-of-bounds kernel write that corrupts adjacent allocations and panics the kernel. Reproduced on 7.0.0-rc7 with UML + KASAN via a crafted image and a single 'touch' inside the mounted directory; crash site resolves to fs/ntfs3/index.c at the memmove. Trigger requires only local mount of an attacker-supplied filesystem image (USB, loopback, or removable media auto-mount). Reject the split whenever the chosen sp plus its declared size already extends past hdr1->used. This is the minimal fix; it preserves the existing hdr_find_split() contract and relies on the same out: cleanup path as the pre-existing error returns. A prior OOB read in the very same indx_insert_into_buffer() memmove was fixed in commit b8c44949044e ("fs/ntfs3: Fix OOB read in indx_insert_into_buffer") by tightening hdr_find_e(), but that fix does not cover the split-point size field path addressed here: sp is returned by hdr_find_split(), not hdr_find_e(), and the underflow is driven by sp->size rather than hdr->used exceeding hdr->total. | 16 Oct |
| CVE-2026-45311 | 9.6 critical | 1.0% | no CPE yet | CodeWhale is a DeepSeek + MiMo coding agent in terminal. From 0.3.0 to 0.8.23, the run_tests tool executes cargo test in the workspace with ApprovalRequirement::Auto, meaning it runs without any user approval prompt. cargo test compiles and executes arbitrary code: test binaries, build.rs build scripts, and proc macros. While auto-approving test execution is a deliberate design choice, it creates an inconsistency in the security boundary. However, in a malicious repository, test code can execute arbitrary shell commands, exfiltrate credentials, or establish persistence with zero approval. The attack is amplified by AGENTS.md (auto-loaded into the system prompt), which can instruct the model to run tests proactively at session start. This vulnerability is fixed in 0.8.23. | Mon 26 Oct |
| CVE-2026-65605 | 9.4 critical | 1.0% | no CPE yet | SiYuan before v3.7.2 contains a stored cross-site scripting vulnerability in Attribute View (database) cell rendering. A Template column value is rendered as HTML via text/template without auto-escaping, and EscapeHTML is only applied when HasUnclosedHtmlTag returns true; because balanced self-closing tags such as <img> are skipped by that check, a payload like <img src=x onerror=...> is stored unescaped and later inserted into the page via innerHTML, executing when the database is viewed. Because the desktop renderer runs with nodeIntegration enabled, the injected script can reach require and escalate to arbitrary command execution. | Sat 24 Oct |
| CVE-2026-70558 | 9.3 critical | 1.0% | no CPE yet | Dinky's POST /download/uploadFromRsByLocal handler passes the caller-supplied path parameter directly to new File(path) and file.transferTo(dest) with no path validation. The route is marked @SaIgnore and /download/** is excluded from the Sa-Token interceptor, so the only guard is a header equality check against a dinkyToken value whose default (efda1551-7958-4e0f-80a8-dfd107df3e38) is hardcoded in source and shipped to every deployment. Anyone who can reach Dinky's HTTP port (8888 by default) and supplies the hardcoded token can write arbitrary files as the Dinky service account. The default Docker image runs on 8888 with no proxy or authentication and chmod 777 on /opt/dinky, so the application's own classpath, launch scripts, and static assets are writable. Demonstrated impact: overwriting /opt/dinky/config/static/index.html served attacker JavaScript to admin browsers immediately, and writing /opt/dinky/org/dinky/Dinky.class executed attacker code as the Dinky service account at the next JVM start via a classpath-shadow launched by script/bin/auto.sh. Writes are uid 9999 (flink), not root, so /etc, /root, /home, and /usr are refused. Affects Dinky v1.2.5 (the current release) and the development branch, where the code is byte-identical. | Sat 24 Oct |
| CVE-2026-103091 | 9.3 critical | 1.0% | no CPE yet | Missing Authorization (CWE-862) and Initialization of a Resource with an Insecure Default (CWE-1188) in the MCP proxy (HandleStreamPassThrough in mcp_service.go) and OIDC registration flow (GET /info and POST /register in internal/controlplane/server.go) in Google OSS Agent Mesh (SAM) versions 0.1.0-alpha.1 through 0.1.0-alpha.8 allow a remote attacker to self-enroll into the mesh and invoke arbitrary tools across all connected MCP backend nodes. Because GET /info publicly discloses OIDC configuration, POST /register auto-enrolls any valid OIDC token holder without claim filtering or ban checks, and the default policy binds sam:system:authenticated to wildcard services and targets, an attacker can obtain a mesh credential with full read, write, and execution access on backend services. Users should upgrade to version 0.1.0-alpha.9 or later, which enforces a fail-closed default policy, explicit policy-binding requirements on enrollment, and ban checks. | 19 Oct |
| CVE-2025-34328 | 9.3 critical | 1.0% | audiocodes, fax_server, interactive_voice_response | AudioCodes Fax Server and Auto-Attendant IVR appliances versions up to and including 2.6.23 include a web administration component (F2MAdmin) that exposes an unauthenticated script-management endpoint at AudioCodes_files/utils/IVR/diagram/ajaxScript.php. The saveScript action writes attacker-supplied data directly to a server-side file path under the privileges of the web service account, which runs as NT AUTHORITY\\SYSTEM on Windows deployments. A remote, unauthenticated attacker can write arbitrary files into the product’s web-accessible directory structure and subsequently execute them. | 18 Oct |
| CVE-2025-34329 | 9.3 critical | 1.0% | audiocodes, fax_server, interactive_voice_response | AudioCodes Fax Server and Auto-Attendant IVR appliances versions up to and including 2.6.23 expose an unauthenticated backup upload endpoint at AudioCodes_files/ajaxBackupUploadFile.php in the F2MAdmin web interface. The script derives a backup folder path from application configuration, creates the directory if it does not exist, and then moves an uploaded file to that location using the attacker-controlled filename, without any authentication, authorization, or file-type validation. On default Windows deployments where the backup directory resolves to the system drive, a remote attacker can upload web server or interpreter configuration files that cause a log file or other server-controlled resource to be treated as executable code. This allows subsequent HTTP requests to trigger arbitrary command execution under the web server account, which runs as NT AUTHORITY\\SYSTEM. | 18 Oct |
| CVE-2025-66448 | 8.8 high | 1.0% | vllm | vLLM is an inference and serving engine for large language models (LLMs). Prior to 0.11.1, vllm has a critical remote code execution vector in a config class named Nemotron_Nano_VL_Config. When vllm loads a model config that contains an auto_map entry, the config class resolves that mapping with get_class_from_dynamic_module(...) and immediately instantiates the returned class. This fetches and executes Python from the remote repository referenced in the auto_map string. Crucially, this happens even when the caller explicitly sets trust_remote_code=False in vllm.transformers_utils.config.get_config. In practice, an attacker can publish a benign-looking frontend repo whose config.json points via auto_map to a separate malicious backend repo; loading the frontend will silently run the backend’s code on the victim host. This vulnerability is fixed in 0.11.1. | 18 Oct |
| CVE-2026-7607 | 8.7 high | 1.0% | trendnet, tew-821dap_firmware | A security vulnerability has been detected in TRENDnet TEW-821DAP 1.12B01. Impacted is the function auto_update_firmware of the component Firmware Udpate. The manipulation of the argument str leads to buffer overflow. The attack may be initiated remotely. The vendor explains: "That firmware version will only work on our hardware version v1.xR. We have already EOL that product 8 years ago and are no longer selling". This vulnerability only affects products that are no longer supported by the maintainer. | Mon 26 Oct |
| CVE-2026-58375 | 8.7 high | 1.0% | no CPE yet | JimuReport through 2.5.0 exposes the POST /jmreport/auto/export endpoint without authentication: the handler is annotated @JimuNoLoginRequired, so JimuReportTokenInterceptor skips all authentication and authorization, and the export service streams the rendered report for any supplied report id without verifying the auto-export configuration flag. An unauthenticated remote attacker can enumerate Snowflake report identifiers and export the full contents of any report, including the data returned by the report configured SQL queries and any credentials embedded in its data sources. | Sun 25 Oct |
| CVE-2026-75936 | 8.7 high | 1.0% | no CPE yet | Improper handling of highly compressed data in the GZIP auto-decompression handler in Amazon ion-java before 1.12.0 might allow remote actors to cause a denial of service via a crafted compressed Ion document that expands to an arbitrarily large size upon decompression. To remediate this issue, users should upgrade to version 1.12.0 and configure withGzipDecompressionEnabled(false) and/or set an explicit withMaximumBufferSize() when parsing untrusted input. | Thu 22 Oct |
| CVE-2025-34331 | 8.7 high | 1.0% | audiocodes, fax_server, interactive_voice_response | AudioCodes Fax Server and Auto-Attendant IVR appliances versions up to and including 2.6.23 contain an unauthenticated file read vulnerability via the download.php script. The endpoint exposes a file download mechanism that lacks access control, allowing remote, unauthenticated users to request files stored on the appliance based solely on attacker-supplied path and filename parameters. While limited to specific file extensions permitted by the application logic, sensitive backup archives can be retrieved, exposing internal databases and credential hashes. Successful exploitation may lead to disclosure of administrative password hashes and other sensitive configuration data. | 18 Oct |
| CVE-2026-102028 | 8.7 high | 1.0% | no CPE yet | A NULL Pointer Dereference vulnerability in the routing protocol daemon (rpd) of Juniper Networks Junos OS and Junos OS Evolved allows a remote BGP peer to cause a denial of service when multiple UPDATEs occur over a single TCP session with 'bgp family l2vpn auto-discovery-only' configured. | 16 Oct |
| CVE-2026-76841 | 8.7 high | 1.0% | no CPE yet | Xinference loads models with Hugging Face remote code execution unconditionally enabled, and before version 2.12.0 exposes no setting to disable it. Six loader call sites pass trust_remote_code=True as a literal or as an unconditional default: RerankModel._get_tokenizer in xinference/model/rerank/core.py, SentenceTransformerRerankModel.load in xinference/model/rerank/sentence_transformers/core.py, SentenceTransformerEmbeddingModel.load in xinference/model/embedding/sentence_transformers/core.py, FlagEmbeddingModel.load in xinference/model/embedding/flag/core.py, and two sites in xinference/model/llm/transformers/core.py where PytorchModel._sanitize_model_config and PytorchModel._get_components default the value to True. Because a caller with model launch access can register a model whose type is unknown and supply an arbitrary model path, the server reaches _auto_detect_type and then AutoTokenizer.from_pretrained, which imports and executes Python declared by the model directory's own tokenizer_config.json auto_map, running attacker-supplied code with the privileges of the worker process. Version 2.12.0 gates every site behind allow_trust_remote_code and the XINFERENCE_TRUST_REMOTE_CODE setting, permitting remote code only for bundled built-in models. | 12 Oct |
| CVE-2026-56075 | 8.7 high | 1.0% | no CPE yet | PraisonAI before 4.5.128 contains an arbitrary shell command execution vulnerability where the UI modules hardcode approval_mode to auto, overriding administrator configuration from PRAISON_APPROVAL_MODE environment variable. Authenticated attackers can instruct the LLM agent to execute arbitrary shell commands via subprocess.run with shell=True, bypassing the manual approval gate and insufficient command sanitization blocklists. | 8 Oct |
| CVE-2026-15343 | 8.6 high | 1.0% | no CPE yet | A path traversal vulnerability was identified in GitHub Enterprise Server that allowed an attacker who had code execution inside the Dependabot updater container to write files to arbitrary repository paths, including GitHub Actions workflow files under .github/workflows/ as the path validation did not check the effective path which the attacker could control through the dependency file's directory and symlink target. If the repository used a pull_request_target workflow or had auto-merge enabled, an injected workflow could execute with access to the repository's GitHub Actions secrets. This vulnerability affected all versions of GitHub Enterprise Server prior to 3.22 and was fixed in versions 3.21.3, 3.20.5, 3.19.9, 3.18.12, 3.17.18. | Sat 24 Oct |
| CVE-2026-10591 | 8.6 high | 1.0% | amazon, kiro_ide | Insufficient access control restrictions in the file write tool in Amazon Kiro IDE before version 0.11 might allow remote unauthenticated actors to execute arbitrary commands via crafted instructions that cause writes to execution-sensitive paths (such as .vscode/tasks.json), enabling auto-execution on folder open. To remediate this issue, users should upgrade to Kiro IDE version 0.11 or later. | 20 Oct |
| CVE-2026-97200 | 8.2 high | 1.0% | no CPE yet | Allocation of Resources Without Limits or Throttling vulnerability in hackney_conn in benoitc hackney allows a malicious HTTP server to exhaust client memory via an oversized or highly compressed response body. hackney:request/5 and hackney:body/1 return the whole response body, and nothing bounds how much the connection process buffers. Any server the client requests can trigger it. hackney_conn:read_full_body/2 and the HTTP/2 and HTTP/3 accumulators append every chunk until the body ends. With {auto_decompress, true} the buffered body is then inflated in one step with zlib:gunzip/1 or zlib:inflate/2, which return the whole output at once. This issue affects hackney: from 2.0.0-beta.1 before 4.9.0. | Mon 26 Oct |
| CVE-2026-41113 | 8.1 high | 1.0% | no CPE yet | sagredo qmail before 2026.04.07 allows tls_quit remote code execution because of popen in notlshosts_auto in qmail-remote.c. | Mon 26 Oct |
| CVE-2026-39861 | 7.7 high | 1.0% | anthropic, claude_code | Claude Code is an agentic coding tool. Prior to version 2.1.64, Claude Code's sandbox did not prevent sandboxed processes from creating symlinks pointing to locations outside the workspace. When Claude Code subsequently wrote to a path within such a symlink, its unsandboxed process followed the symlink and wrote to the target location outside the workspace without prompting the user for confirmation. This allowed a sandbox escape where neither the sandboxed command nor the unsandboxed app could independently write outside the workspace, but their combination could write to arbitrary locations, potentially leading to code execution outside the sandbox. Reliably exploiting this required the ability to add untrusted content into a Claude Code context window to trigger sandboxed code execution via prompt injection. Users on standard Claude Code auto-update have received this fix automatically. Users performing manual updates are advised to update to version 2.1.64 or later. | Mon 26 Oct |
| CVE-2026-110205 | 7.7 high | 1.0% | roocode, roo_code | Roo Code through 3.54.0 contains a command approval bypass vulnerability in parseCommandLine() that discards command substitutions placed next to * or ? glob characters. Attackers using indirect prompt injection can craft commands like echo *$(...) that match allowlisted prefixes, executing denied commands without user approval when auto-approval is enabled. | Mon 26 Oct |
| CVE-2026-113352 | 7.7 high | 1.0% | no CPE yet | uTox through 0.18.1 contains a path traversal vulnerability that allows Tox friends to write arbitrary files by sending absolute or ../ file names to victims with auto-accept enabled. Attackers can overwrite shell startup files or plant crafted .ftinfo resume records that ft_find_resumeable() loads, abusing FILE pointers to write /proc/self/mem and execute code. | 20 Oct |
| CVE-2026-110206 | 7.7 high | 1.0% | roocode, roo_code | Roo Code through 3.54.0 contains a command approval bypass vulnerability in parseCommandLine() that allows attackers to auto-approve arbitrary shell commands by placing a command substitution in command position. Attackers using indirect prompt injection can submit commands like $(echo touch /tmp/x) that pass the echo allowlist check, causing the shell to execute the output without user approval. | 16 Oct |
| CVE-2026-110203 | 7.7 high | 1.0% | roocode, roo_code | Roo Code through 3.54.0 contains a command approval bypass vulnerability that allows attackers to auto-approve denied shell commands by abusing quote-masking regex in parseCommandLine(). Attackers can use indirect prompt injection with escaped or single-quoted double quotes to hide command separators behind allowlisted prefixes like echo, executing commands without user approval. | 15 Oct |
50 of 273 · Next
Download these 273 as CSV · subscribe to this list.
EPSS scores: 2026-10-28. NVD cursor: 2026-10-29 00:22. The News page shows only notable CVEs (KEV, your stack, new criticals, EPSS over 50%, or active exploitation per SSVC); this table has all of them.