RAGFlow is an open-source RAG (Retrieval-Augmented Generation) engine. In 0.24.0 and earlier, a Jinja2 template injection in the prompt generator (rag/prompts/generator.py) allows any authenticated user to execute arbitrary OS commands on the server. Any normal user can register, create a Canvas workflow with a DuckDuckGo + LLM component chain, and trigger the SSTI.

RAGFlow through 0.18.1 allows account takeover because it is possible to conduct successful brute-force attacks against email verification codes to perform arbitrary account registration, login, and password reset. Codes are six digits and there is no rate limiting.

RAGFlow is an open-source RAG (Retrieval-Augmented Generation) engine. Versions 0.15.1 and prior are vulnerable to SQL injection. The ExeSQL component extracts the SQL statement from the input and sends it directly to the database query. As of time of publication, no patched version is available.

RAGFlow is an open-source RAG (Retrieval-Augmented Generation) engine. In versions 0.24.0 and prior, a Server-Side Template Injection (SSTI) vulnerability exists in RAGFlow's Agent workflow Text Processing (StringTransform) and Message components. These components use Python's jinja2.Template (unsandboxed) to render user-supplied templates, allowing any authenticated user to execute arbitrary operating system commands on the server. At time of publication, there are no publicly available patches.

RAGFlow before 0.26.3 contains a server-side request forgery vulnerability in the agent workflow "Invoke" component (agent/component/invoke.py). The component builds an outbound request URL from canvas configuration and runtime template variables and passes it to requests.get, requests.post, or requests.put without calling the shared asserturlissafe validator or pinning the resolved address, unlike the crawler, SearXNG, file-upload, and RSS fetch paths. A user who can create or trigger an agent can direct the server to fetch loopback, link-local, and RFC 1918 destinations, including cloud instance metadata endpoints and services co-located on the deployment network, and the response body is returned as the component output. Where an agent is configured to interpolate the chat query into the Invoke URL, the destination is chosen by whoever can send that query.

RAGFlow is an open-source RAG (Retrieval-Augmented Generation) engine based on deep document understanding. An authenticated user can exploit the Insecure Direct Object Reference (IDOR) vulnerability that may lead to unauthorized cross-tenant access (list tenant user accounts, add user account into other tenant). Unauthorized cross-tenant access: list user from other tenant (e.g., via GET /<tenantid>/user/list), add user account to other tenant (POST /<tenantid>/user). This issue has not yet been patched. Users are advised to reach out to the project maintainers to coordinate a fix.

RAGFlow 0.13.0 suffers from improper access control in document-hooks.ts, allowing unauthorized access to user documents.

Stored Cross-site Scripting (XSS) vulnerability in api.apps.dialogapp.setdialog in RAGFlow 0.17.2 allows remote attackers to execute arbitrary JavaScript via crafted input to the assistant greeting field, which is stored unsanitised and rendered using a markdown component with rehype-raw.