4084 matches found
Unsafe Deserialization in Public `keras.layers.TorchModuleWrapper.from_config` Reaches `torch.load(weights_only=False)` Without Explicit Unsafe Opt-In
Description keras.layers.TorchModuleWrapper.fromconfig can deserialize attacker-controlled PyTorch pickle data through torch.load..., weightsonly=False when it is called directly through the public Keras layer API. TorchModuleWrapper is exported as a public Keras layer:...
SSRF via Unvalidated URL in Artefact Import and Datastore Scrape (Incomplete CVE-2026-0560 Fix)
Description The fix for CVE-2026-0560 commit 76a54f0 added validateurl SSRF protection to backend/routers/files.py only. Two other endpoints that fetch user-supplied URLs have no SSRF protection and pass attacker-controlled URLs directly to ScrapeMaster: 1. POST...
Pipelines Frontend /_proxy/ Unrestricted SSRF
This report is not public...
TFSMLayer bypasses `safe_mode=True`, allowing attacker-controlled code execution during model inference
Summary TFSMLayer allows loading attacker-controlled TensorFlow SavedModels when deserializing a .keras model, even when safemode=True the default. While TensorFlow does not execute SavedModel functions during load, the attacker-controlled graph is registered during deserialization and executes...
Integer Overflow → Heap Buffer Overflow in BYTES-Tensor Parsing (DoS)
This report is not public...
LangChain HTMLSectionSplitter – XXE caused by unsafe XSLT parsing
This report is not public...
Path Traversal in Tokenizer Conversion Script
The script for converting slow tokenizers is vulnerable to a Path Traversal attack via the --checkpointname command-line argument. This allows an attacker to create files outside of the intended dumppath directory. Vulnerable Code Location: The vulnerability is located in the logic for converting...
Denial of Service(DOS) in JSONReader
Description There exists a denial of service vulnerabilityDOS that occurs by python hitting max recursion depth while parsing a deeply nested json file using JSONReader. Vulnerable piece of code...
Regular expression Denial of Service - ReDoS in huggingface/transformers
Description A regular expression denial of service ReDoS vulnerability has been identified in the Hugging Face Transformers library's Donut processor. The vulnerability exists in the token2json method of the DonutProcessor class, which processes document tokens into JSON format. The regex pattern...
Arbitary file read through path traversal
Description: The code in genericutils.py has a path traversal vulnerability, which allows an attacker to control the file path provided to the ImageDocument class. This can lead to the reading of arbitrary files on the server, including sensitive system files, through base64 encoding and decoding...
XSS vulnerability exists in some specific browsers
Description The XSS vulnerability cannot be triggered in Chrome, but it is triggered when using Firefox and the latest version of Firefox. Since Firefox is widely used, when the administrator uses Firefox to view the relevant interface, the XSS vulnerability will be triggered, resulting in the...
Using Mermaid to cause JS memory overflow and service downtime
Description Librechat has many means of limiting the rate, which can be found at https://www.librechat.ai/docs/configuration/librechatyaml/objectstructure/configratelimits. However, it can be found that the Fork Function in /api/convos/fork is not restricted, which allows attackers to fork...
Path Traversal via Symbolic Links in `ObsidianReader`
Description The ObsidianReader class, designed to parse Obsidian vaults, contains a critical security flaw that allows arbitrary file read through symbolic links symlinks. When processing a vault, the reader does not resolve or validate the absolute paths of files, enabling an attacker to place a...
SQL injection vulnerabilities in multiple vector stores
Description Multiple vector store integrations have SQL injection vulnerabilities, which can allow an attacker to read and write data using SQL. Example vulnerable code snippet in the Couchbase vector store integration: def deleteself, refdocid: str, kwargs: Any - None: """ Delete a document by i...
Bug Bounty Report: Command Injection Vulnerability in subprocess Call
This report is not public...
AutoGPT SSTI Vulnerability Leading to Remote Code Execution (RCE)
Summary AutoGPT, an open-source AI tool that automates task execution, is vulnerable to a Server-Side Template Injection SSTI that could lead to arbitrary command execution. The vulnerability arises from the improper handling of user-supplied format strings in the AgentOutputBlock implementation,...
A SQL Injection in DuckDB via prompt can lead to RCE
Target Link Description sql = f""" SELECT ftsmainself.tablename.matchbm25self.nodeidcolumn, 'query' AS score, self.nodeidcolumn, self.textcolumn FROM self.tablename WHERE score IS NOT NULL ORDER BY score DESC LIMIT self.similaritytopk; """ The duckdbretriever performs "search using string" and...
Stored Cross-Site Scripting (XSS) via SAML IdP XML Injection
An attacker can achieve stored cross-site scripting XSS by injecting malicious JavaScript into the SAML IdP XML metadata. This metadata is used to generate the SAML login redirect URL, which is ultimately set as the value of window.location.href. This vulnerability allows the attacker to execute...
malicious gguf model can cause DoS by allocate unlimited memory via network access
This report is not public...
SSRF due to insufficient patch of CVE-2024-5822
This report is not public...
SSRF Vulnerabilities found in Search and Github Integration AutoGPT Blocks
Hi, AutoGPT developers! Summary I have identified several Server-Side Request Forgery SSRF vulnerabilities in the default agent blocks provided by the AutoGPT platform. These vulnerabilities could lead to severe security issues, including credential leakage e.g., GitHub tokens, internal network...
Logs Debug Injection In File Download
Description In 2 API: /code/download/:sessionId/:fileId and /download/:userId/:fileid The parameters sessionId, fileId, userId, fileid are not validated or filtered at all but are saved directly to log.debug Proof of Concept Prepare: The logs file on the server is stored at /app/api/debug-.log I...
SSRF via POST /v1/llm/add_llm and /v1/conversation/tts
This report is not public...
Improper Role Modification by Admins for Billing Permissions
Description Admins, who do not have direct permissions to access billing resources, are able to change the permissions of existing users to have billing permissions. This can lead to a privilege escalation scenario where an administrator can: 1. Change the role of an existing user to include...
Lack of access control on /users/me/org endpoint
Description The /users/me/org route is not adequately protected by access control mechanisms such as a middleware. This lack of authorization allows unauthorized users to access information about all team members in the current organization, even if the user does not have sufficient privileges. A...
Denial of service through sshfs-client in tracking server
This report is not public...
Running user provided regular expression, lead to DOS
This report is not public...
Open Redirect
This report is not public...
Redos (Regular Expression Denial of Service)
This report is not public...
server crash by zip bomb
This report is not public...
Store XSS when Edit label set
Description Store XSS when Edit label set. I noticed, you have filtered the input when creating the label set. But, perhaps you forgot to filter when editing the label set. Proof of Concept 1 .Create a label set 2 .Edit label set with payload : haidoalertdocument.domain 3 .Click Export multiple...
Add arbitrary users to the user group
Description Add arbitrary users to the user group Proof of Concept 1 .Administrator user haido456 creates a user group name : group456 2 .User hai123 has general user rights but has the right to add arbitrary users to the user group: group456 3 .This includes users that the admin does not want...
Important Cookie without Secure flag
Description Cookie accessToken is without Secure flag. Mentioned cookie is responsible for user auth. Proof of Concept Repro steps: As logged in user https://app.vrite.io/ open DevTools and check Cookies table, get value of accessToken cookie. Open other browser, go to app.vrite.io site, open...
Insufficient Session Expiration
Insufficient session expiration is a web application security vulnerability that occurs when a web application does not properly manage the lifecycle of a user's session. This can allow an attacker to hijack the user's session and gain unauthorized access to the application. The web application m...
Cross-Site Request Forgery (CSRF)
A Cross-site request forgery CSRF attack is a type of malicious attack whereby an attacker tricks a victim into performing an action on a website that they do not intend to do. This can be done by sending the victim a malicious link or by exploiting a vulnerability in the website. For example, an...
stored XSS Bypass in the TAGS Section and other places in the application
Hello, I was able to bypass the XSS Protection and get a stored XSS using the XSS Payload in the Video and Screenshots. Thank you for your time and effort. Best regards Ahmed Hassan...
Path Traversal in uploadAttachment
POC : see https://1drv.ms/v/s!Avwg5C1eKVA4gl3LF2hgRyVNrSqk?e=DHbHKF We also contact the Maintainer through email lujie.ac.cn...
Dos via Document Comments
Description An attacker can abuse the document comment functionality, handled by the /api/comments.create API endpoint, since there is not size check or validation of the comment contents, which allows an attacker to send a comment with almost an unlimited number of characters1MB max POST size...
Broken Authentication
Description I tested the demo site you provided. I see that there is an Broken Authentication vulnerability in Administration: CPU stats API. The Administration: CPU stats API does not validated user permissions. Proof of Concept link video PoC https://screenpal.com/watch/c01F1bVBmX1 Step 1. In t...
Stored XSS on user "Edit own profile" function
Description An attacker can inject malicious executable scripts into the code of the Social media field Proof of Concept Log in as a Member user, access My profile - Edit own profile function, insert this payload to any field " autofocus onfocus=promptdocument.domain then click Save. Access the...
The user can delete himself
Description Bypassing the conditional check leads to the user can delete himself. Proof of Concept Step 1: The user with id 18834 attempts to delete himself but encounter an error Step 2: By using userid=18834' instead of userid=18834, the user was able to successfully delete himself...
Sensitive Cookie Without HttpOnly Flag
Description Access and login to the demo website: https://demo.openitcockpit.io/ Press F12 on your keyboard or right-click on the website to open dev-tool. At Application tab, choose Cookies and there is CookieAuth sensitive cookie without HttpOnly flag. Proof of Concept Link image evidence:...
Reflected XSS in Path Traversal detector
Description Azuracast has a feature that block all Path Traversal tentative good job implementing it. But when azuracast block an attack reflect the path without sanitize the output PathTraversalDetected.php. It is possibile to do attack like Reflected XSS or HTML injection. Step to reproduce 1. ...
XSS Stored in Caption Image
Description Hello team, I found an xss stored in the caption field as demonstrated in the gif below. Proof of Concept...
Cross Site Scripting (XSS) in UrlSlug
Description Please enter a description of the vulnerability. Cross Site Scripting XSS in UrlSlug of pimcore/pimcore Its Different than https://huntr.dev/bounties/75bc7d07-46a7-4ed9-a405-af4fc47fb422/ Proof of Concept 1. Login in stable account URL : https://11.x-dev.pimcore.fun/admin/ 2. Go to...
Path Traversal in code
Description The product uses external input to construct a pathname that should be within a restricted directory, but it does not properly neutralize '.../...//' doubled triple dot slash sequences that can resolve to a location that is outside of that directory. Proof of Concept Code that has the...
Simple graph bed system has deserialization vulnerability and weak type comparison vulnerability
Description Simple graph bed has deserialization vulnerability and weak type comparison vulnerability Proof of Concept As you can see on line 129 below, there is a deserialization point and it is cookie passed The user controlled auth complex value in the cookie is given to the browsercookie...
Anti-CSRF mechanism is not present
Description The application is vulnerable to a CSRF attack. Proof of Concept 1. Login as admin. 2. Open the following HTML file in the browser. This action is equivalent to clicking a link sent by an attacker. trap.html html history.pushState'', '', '/' 3. Click the button. 4. A new user is creat...
Cookie without “Secure “ and “ HttpOnly ” flag attribute
Description HttpOnly and Secure attribute is not set for session cookies in the application. Proof of Concept https://drive.google.com/file/d/1ZAanmAbOn-jSf6ZMS5JIQKUzJ78fUrea/view?usp=sharing...
XSS via markdown syntax
Description Hi,Maintainer,thanks for reading.I am glad to report a secure problem to you. I found that your forum allows users to use markdown syntax to post articles and comments, but there is no corresponding protection means, which is unsafe. Any user can post dangerous content, like the...