LibreChat is a ChatGPT clone with additional features. From 0.8.2 to 0.8.2-rc3, The MCP (Model Context Protocol) OAuth callback endpoint accepts the redirect from the identity provider and stores OAuth tokens for the user who initiated the flow, without verifying that the browser hitting the redirect URL is logged in or that the logged-in user matches the initiator. An attacker can send the authorization URL to a victim; when the victim completes the flow, the victim’s OAuth tokens are stored on the attacker’s LibreChat account, enabling account takeover of the victim’s MCP-linked services (e.g. Atlassian, Outlook). This vulnerability is fixed in 0.8.3-rc1.
A vulnerability in danny-avila/librechat allows attackers to exploit the unrestricted Fork Function in `/api/convos/fork` to fork numerous contents rapidly. If the forked content includes a Mermaid graph with a large number of nodes, it can lead to a JavaScript heap out of memory error upon service restart, causing a denial of service. This issue affects the latest version of the product.
LibreChat is a ChatGPT clone with additional features. Prior to v0.8.2-rc2, LibreChat's MCP stdio transport accepts arbitrary commands without validation, allowing any authenticated user to execute shell commands as root inside the container through a single API request. This vulnerability is fixed in v0.8.2-rc2.
LibreChat is a ChatGPT clone with additional features. Version 0.8.1-rc2 is prone to a server-side request forgery (SSRF)
vulnerability due to missing restrictions of the Actions feature in the default configuration. LibreChat enables users to configure agents with predefined instructions and actions that can interact with remote services via OpenAPI specifications, supporting various HTTP methods, parameters, and authentication methods including custom headers. By default, there are no restrictions on accessible services, which means agents can also access internal components like the RAG API included in the default Docker Compose setup. This issue is fixed in version 0.8.1-rc2.
LibreChat is a ChatGPT clone with additional features. Version 0.8.1-rc2 does not enforce proper access control when
querying agent permissions. An authenticated attacker can read the permissions of arbitrary agents, even if they have no permissions for this agent. LibreChat allows the configuration of agents that have a predefined set of instructions and context. Private agents are not visible to other users. However, if an attacker knows the agent ID, they can read the permissions of the agent including the permissions individually assigned to other users. This issue is fixed in version 0.8.2-rc2.
LibreChat is a ChatGPT clone with additional features. Version 0.8.1-rc2 does not enforce proper access control for file uploads to an agents file context and file search. An authenticated attacker with access to the agent ID can change the behavior of arbitrary agents by uploading new files to the file context or file search, even if they have no permissions for this agent. This issue is fixed in version 0.8.2-rc2.
LibreChat is a ChatGPT clone with additional features. In versions 0.8.0 and below, there is no handler for JSON parsing errors; SyntaxError from express.json() includes user input in the error message, which gets reflected in responses. User input (including HTML/JavaScript) can be exposed in error responses, creating an XSS risk if Content-Type isn't strictly enforced. This issue does not have a fix at the time of publication.
LibreChat is a ChatGPT clone with additional features. In versions 0.8.0 and below, when creating prompts, JSON requests are sent to define and modify the prompts via PATCH endpoint for prompt groups (/api/prompts/groups/:groupId). However, the request bodies are not sufficiently validated for proper input, enabling users to modify prompts in a way that was not intended as part of the front end system. The patchPromptGroup function passes req.body directly to updatePromptGroup() without filtering sensitive fields. This issue is fixed in version 0.8.1.
LibreChat is a ChatGPT clone with additional features. In versions 0.8.0 and below, when a user posts a question, the iconURL parameter of the POST request can be modified by an attacker. The malicious code is then stored in the chat which can then be shared to other users. When sharing chats with a potentially malicious “tracker”, resources loaded can lead to loss of privacy for users who view the chat link that is sent to them. This issue is fixed in version 0.8.1.
LibreChat is a ChatGPT clone with additional features. Prior to version 0.8.1-rc2, LibreChat is vulnerable to Server-side Request Forgery (SSRF), by passing specially crafted OpenAPI specs to its "Actions" feature and making the LLM use those actions. It could be used by an authenticated user with access to this feature to access URLs only accessible to the LibreChat server (such as cloud metadata services, through which impersonation of the server might be possible). This issue has been patched in version 0.8.1-rc2.
LibreChat version 0.7.9 is vulnerable to a Denial of Service (DoS) attack due to unbounded parameter values in the `/api/memories` endpoint. The `key` and `value` parameters accept arbitrarily large inputs without proper validation, leading to a null pointer error in the Rust-based backend when excessively large values are submitted. This results in the inability to create new memories, impacting the stability of the service.
In danny-avila/librechat version 0.7.9, there is an insecure API design issue in the 2-Factor Authentication (2FA) flow. The system allows users to disable 2FA without requiring a valid OTP or backup code, bypassing the intended verification process. This vulnerability occurs because the backend does not properly validate the OTP or backup code when the API endpoint '/api/auth/2fa/disable' is directly accessed. This flaw can be exploited by authenticated users to weaken the security of their own accounts, although it does not lead to full account compromise.
A vulnerability in danny-avila/librechat version 0.7.9 allows for HTML injection via the Accept-Language header. When a logged-in user sends an HTTP GET request with a crafted Accept-Language header, arbitrary HTML can be injected into the <html lang=""> tag of the response. This can lead to potential security risks such as cross-site scripting (XSS) attacks.
A mass assignment vulnerability exists in danny-avila/librechat, affecting all versions. This vulnerability allows attackers to manipulate sensitive fields by automatically binding user-provided data to internal object properties or database fields without proper filtering. As a result, any extra fields in the request body are included in agentData and passed to the database layer, allowing overwriting of any field in the schema, such as author, access_level, isCollaborative, and projectIds. Additionally, the Object.Prototype can be polluted due to the use of Object.assign with spread operators.
danny-avila/librechat is affected by an authorization bypass vulnerability due to improper access control checks. The `checkAccess` function in `api/server/middleware/roles/access.js` uses `permissions.some()` to validate permissions, which incorrectly grants access if only one of multiple required permissions is present. This allows users with the 'USER' role to create agents despite having `CREATE: false` permission, as the check for `['USE', 'CREATE']` passes with just `USE: true`. This vulnerability affects other permission checks as well, such as `PROMPTS`. The issue is present in all versions prior to the fix.
In version 0.7.8 of danny-avila/librechat, improper authorization controls in the conversation sharing feature allow unauthorized access to other users' conversations if the conversation ID is known. Although UUIDv4 conversation IDs are generated server-side and are difficult to brute force, they can be obtained from less-protected sources such as server-side access logs, browser history, or screenshots. The vulnerability permits a logged-in user to gain read-only access to another user's conversations by exploiting the `/api/share/conversationID` endpoint, which lacks authorization checks. This issue is resolved in version v0.7.9-rc1.
LibreChat is a ChatGPT clone with additional features. In versions 0.0.6 through 0.7.7-rc1, an exposed testing endpoint allows reading arbitrary chats directly from the Meilisearch engine. The endpoint /api/search/test allows for direct access to stored chats in the Meilisearch engine without proper access control. This results in the ability to read chats from arbitrary users. This issue is fixed in version 0.7.7.
In version 0.7.5 of danny-avila/LibreChat, there is an improper access control vulnerability. Users can share, use, and create prompts without being granted permission by the admin. This can break application logic and permissions, allowing unauthorized actions.
In danny-avila/librechat version git 0c2a583, there is an improper input validation vulnerability. The application uses multer middleware for handling multipart file uploads. When using in-memory storage (the default setting for multer), there is no limit on the upload file size. This can lead to a server crash due to out-of-memory errors when handling large files. An attacker without any privileges can exploit this vulnerability to cause a complete denial of service. The issue is fixed in version 0.7.6.
A vulnerability in danny-avila/librechat version git a1647d7 allows an unauthenticated attacker to cause a denial of service by sending a crafted payload to the server. The middleware `checkBan` is not surrounded by a try-catch block, and an unhandled exception will cause the server to crash. This issue is fixed in version 0.7.6.