| CVE |
Vendors |
Products |
Updated |
CVSS v3.1 |
| Langflow is a tool for building and deploying AI-powered agents and workflows. Prior to 1.9.2, by controlling a files that are digested into the RAG, an attacker can direct the node to read any file on the file-system by absolute path. All components based on BaseFileComponent are vulnerable to the vulnerability. This includes Docling (DoclingInlineComponent), Docling Serve, DoclingRemoteComponent), Read File (FileComponent), NVIDIA Retriever Extraction (NvidiaIngestComponent), Video File (VideoFileComponent), and Unstructured API (UnstructuredComponent). This vulnerability is fixed in 1.9.2. |
| vLLM versions >= 0.10.2 and < 0.13.0 are missing sparse tensor validation in multimodal embeddings processing. Because PyTorch disables sparse tensor invariant checks by default, an attacker can submit crafted embedding requests with malformed (negative or out-of-bounds) tensor indices, when the prompt-embeds feature is enabled, to trigger crashes or resource exhaustion (denial of service), with potential for out-of-bounds/write-what-where memory corruption. This continues CVE-2025-62164, whose prior fix only disabled the feature by default rather than addressing the root cause. |
| Capgo before 12.128.2 allows non-admin API keys to read webhook signing secrets via Supabase REST due to insufficient row-level security policies on the webhooks table. Attackers can retrieve the webhook secret and forge valid X-Capgo-Signature headers to send authenticated webhook events to configured receivers, breaking webhook authenticity and integrity. |
| Capgo before 12.128.2 contains an information disclosure vulnerability in the public.exist_app_v2 RPC function that allows unauthenticated attackers to enumerate app_ids by calling POST /rest/v1/rpc/exist_app_v2 with arbitrary appid parameters. Remote attackers can exploit this SECURITY DEFINER function to determine whether specific app_ids exist in the public.apps table, enabling cross-tenant app enumeration and privacy violations. |
| NocoDB is software for building databases as spreadsheets. Prior to 2026.05.1, the shared-view password check fell back to strict-equality (===) comparison for legacy plaintext passwords, leaking the password's length and per-character prefix through response timing. This vulnerability is fixed in 2026.05.1. |
| Unauthenticated credential disclosure in the wizard interface in ZTE ZXHN H188A V6.0.10P2_TE and V6.0.10P3N3_TE allows unauthenticated attackers on the local network to retrieve sensitive credentials from the router's web management interface, including the default administrator password, WLAN PSK, and PPPoE credentials. In some observed cases, configuration changes may also be performed without authentication. |
| Sensitive data exposure leading to admin/WLAN credential leak in ZTE ZXHN H298A 1.1 and H108N 2.6. A crafted request to the router web interface can expose sensitive device and account information. In affected builds, the response may include the administrator password and WLAN PSK, enabling authentication bypass and network compromise. Some firmware versions may expose only partial identifiers (e.g., serial number, ESSID, MAC addresses). |
| An Improper Input Validation in Ivanti EPMM before versions 12.6.1.1, 12.7.0.1, and 12.8.0.1 allows a remotely authenticated user with administrative access to achieve remote code execution. |
| NocoDB is software for building databases as spreadsheets. Prior to 2026.04.1, sign-in response timing differed between known and unknown email addresses because the unknown-user branch returned without performing a password hash comparison. This vulnerability is fixed in 2026.04.1. |
| Hono before 4.12.12 does not validate cookie names on the write path in the setCookie(), serialize(), and serializeSigned() functions, allowing invalid characters such as control characters (e.g. \r or \n) when an application passes a user-controlled cookie name. This can produce malformed Set-Cookie header values. In modern runtimes such as Node.js and Cloudflare Workers, such invalid header values are rejected and cause a runtime error before the response is sent, so header injection or response splitting could not be reproduced; the issue primarily affects correctness and robustness, resulting in runtime errors (availability) rather than confirmed header injection. |
| 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. |
| ACE vulnerability in conditional configuration file processing by QOS.CH logback-core up to and including version 1.5.18 in Java applications, allows an attacker to execute arbitrary code by compromising an existing logback configuration file or by injecting an environment variable before program execution.
A successful attack requires the presence of Janino library and Spring Framework to be present on the user's class path. In addition, the attacker must have write access to a
configuration file. Alternatively, the attacker could inject a malicious
environment variable pointing to a malicious configuration file. In both
cases, the attack requires existing privilege. |
| Caddy is an extensible server platform that uses TLS by default. From 2.7.0 until 2.11.3, the FastCGI transport's splitPos() in modules/caddyhttp/reverseproxy/fastcgi/fastcgi.go misuses golang.org/x/text/search with search.IgnoreCase when the request path contains a non-ASCII byte. Two distinct flaws in that fallback let an attacker mislead Caddy's FastCGI splitting into treating a non-.php (or other configured split_path extension) file as a script. In any deployment where the attacker can place content into a file served via FastCGI (uploads, file storage, etc.), this can be escalated to remote code execution by crafting a URL whose path triggers either flaw. This vulnerability is fixed in 2.11.3. |
| Insertion of Sensitive Information Into Sent Data vulnerability in Stiofan GetPaid allows Retrieve Embedded Sensitive Data.
This issue affects GetPaid: from n/a through 2.8.49. |
| Unauthenticated Sensitive Data Exposure in Amelia <= 2.2 versions. |
| Subscriber Sensitive Data Exposure in Chatway Live Chat – AI Chatbot, Customer Support, FAQ & Helpdesk Customer Service & Chat Buttons <= 1.4.8 versions. |
| Unauthenticated Sensitive Data Exposure in Affiliates Manager <= 2.9.50 versions. |
| Unauthenticated Sensitive Data Exposure in ABC Crypto Checkout <= 1.8.2 versions. |
| Unauthenticated Sensitive Data Exposure in GetGenie <= 4.4.1 versions. |
| The vulnerability is present in the ‘/addJugador’ endpoint:
* The 'keyJugador' and 'keyJugadorObjectiu' parameters allow the modification of other users’ information without requiring prior authorization validation. This could enable an authenticated attacker to alter any user’s ID and change their information.
* The ‘punts’ and ‘numObjectiusEliminats’ fields allow arbitrary data to be added because user input is not properly validated. This makes it possible to obtain authentic prizes, awarded by city councils, by falsifying game scores.
* In the ‘tokens’ field, administrative privileges can be self-assigned without server validation or prior authentication. This vulnerability could allow an authenticated attacker to grant themselves administrator permissions and thus escalate privileges.
* Numeric fields allow the entry of extremely long values, which can cause the system to crash. Successful exploitation of this vulnerability could allow an authenticated attacker to launch a denial-of-service (DoS) attack, preventing created games from being playable.
* The ‘urlImatge’ parameter allows server-side requests to arbitrary URLs, enabling the retrieval of users’ internal IP addresses, access to internal services, reading of local files, and unauthorized interaction with third-party APIs. An authenticated attacker could gain access to sensitive data. |
