An issue in langchain langchain-ai v.0.0.232 and before allows a remote attacker to execute arbitrary code via a crafted script to the PythonAstREPLTool._run component.

An issue in langchain v.0.0.64 allows a remote attacker to execute arbitrary code via the PALChain parameter in the Python exec method.

@langchain/langgraph-checkpoint-redis is the Redis checkpoint and store implementation for LangGraph. A query injection vulnerability exists in the @langchain/langgraph-checkpoint-redis package's filter handling. The RedisSaver and ShallowRedisSaver classes construct RediSearch queries by directly interpolating user-provided filter keys and values without proper escaping. RediSearch has special syntax characters that can modify query behavior, and when user-controlled data contains these characters, the query logic can be manipulated to bypass intended access controls. This vulnerability is fixed in 1.0.2.

In LangChain through 0.0.131, the LLMMathChain chain allows prompt injection attacks that can execute arbitrary code via the Python exec method.

A vulnerability in the GraphCypherQAChain class of langchain-ai/langchain version 0.2.5 allows for SQL injection through prompt injection. This vulnerability can lead to unauthorized data manipulation, data exfiltration, denial of service (DoS) by deleting all data, breaches in multi-tenant security environments, and data integrity issues. Attackers can create, update, or delete nodes and relationships without proper authorization, extract sensitive data, disrupt services, access data across different tenants, and compromise the integrity of the database.

An issue in Harrison Chase langchain v.0.0.194 and before allows a remote attacker to execute arbitrary code via the from_math_prompt and from_colored_object_prompt functions.

In Langchain through 0.0.155, prompt injection allows an attacker to force the service to retrieve data from an arbitrary URL, essentially providing SSRF and potentially injecting content into downstream tasks.

A host header injection vulnerability in SysPass 3.2x allows an attacker to load malicious JS files from an arbitrary domain which would be executed in the victim's browser.

Git LFS is a Git extension for versioning large files. When Git LFS requests credentials from Git for a remote host, it passes portions of the host's URL to the `git-credential(1)` command without checking for embedded line-ending control characters, and then sends any credentials it receives back from the Git credential helper to the remote host. By inserting URL-encoded control characters such as line feed (LF) or carriage return (CR) characters into the URL, an attacker may be able to retrieve a user's Git credentials. This problem exists in all previous versions and is patched in v3.6.1. All users should upgrade to v3.6.1. There are no workarounds known at this time.

HedgeDoc is a platform to write and share markdown. In versions prior to 1.9.0, an unauthenticated attacker can inject arbitrary JavaScript into the speaker-notes of the slide-mode feature by embedding an iframe hosting the malicious code into the slides or by embedding the HedgeDoc instance into another page. The problem is patched in version 1.9.0. There are no known workarounds aside from upgrading.

Smarty is a template engine for PHP, facilitating the separation of presentation (HTML/CSS) from application logic. Prior to versions 3.1.42 and 4.0.2, template authors could run arbitrary PHP code by crafting a malicious math string. If a math string was passed through as user provided data to the math function, external users could run arbitrary PHP code by crafting a malicious math string. Users should upgrade to version 3.1.42 or 4.0.2 to receive a patch.

Improper authorization handling in Zoom Workplace for Android before version 6.5.10 may allow an unauthenticated user to conduct an escalation of privilege via network access.

pyLoad is a free and open-source download manager written in Python. In versions prior to 0.5.0b3.dev91, pyLoad web interface contained insufficient input validation in both the Captcha script endpoint and the Click'N'Load (CNL) Blueprint. This flaw allowed untrusted user input to be processed unsafely, which could be exploited by an attacker to inject arbitrary content into the web UI or manipulate request handling. The vulnerability could lead to client-side code execution (XSS) or other unintended behaviors when a malicious payload is submitted. user-supplied parameters from HTTP requests were not adequately validated or sanitized before being passed into the application logic and response generation. This allowed crafted input to alter the expected execution flow. CNL (Click'N'Load) blueprint exposed unsafe handling of untrusted parameters in HTTP requests. The application did not consistently enforce input validation or encoding, making it possible for an attacker to craft malicious requests. Version 0.5.0b3.dev91 contains a patch for the issue.