Unauthenticated Brute-Force Vulnerability in LightRAG-HKU
The /login endpoint in LightRAG-HKU versions prior to 1.5.5 lacks rate limiting or account lockout, enabling high-speed credential brute-force attacks.
CVE search metadata
CVE search record: CVE-2026-85734. Severity: critical. CVSS: 9.1. KEV: no. Product: lightrag-hku (< 1.5.5). Brief: Unauthenticated Brute-Force Vulnerability in LightRAG-HKU. Brief link: https://feed.craftedsignal.io/briefs/2026-09-lightrag-login-brute-force/
LightRAG-HKU, a framework for knowledge graph retrieval, contains a critical security flaw in its authentication mechanism. The application's /login endpoint, implemented in lightrag/api/lightrag_server.py, fails to implement rate limiting, account lockout, or request delays for failed authentication attempts. This oversight allows an unauthenticated, network-reachable attacker to programmatically iterate through password lists at full network speed to compromise administrative or user accounts. The vulnerability, tracked as CVE-2026-85734, affects all versions prior to 1.5.5. Given the sensitivity of the data stored within LightRAG knowledge graphs, successful exploitation provides unauthorized access to proprietary documents and administrative operations.
Attack Chain
- Attacker performs network reconnaissance to identify reachable instances of LightRAG-HKU on default port 9621.
- Attacker interacts with the target HTTP service to confirm the presence of the
/loginendpoint. - Attacker prepares a dictionary of common passwords or credential lists for brute-force operations.
- Attacker writes a script to automate HTTP POST requests to the
/loginendpoint. - Attacker executes the script, passing user credentials via
form_datawithout encountering server-side throttling. - Attacker monitors the HTTP response codes (e.g., waiting for a 200 OK) to identify successful password matches.
- Attacker uses the compromised credentials to access the LightRAG API and extract sensitive knowledge graph information.
Impact
Successful exploitation allows for full unauthorized access to the LightRAG instance. Depending on the deployment, this could lead to the exposure of confidential knowledge stored in the graph, unauthorized administrative changes, and full exfiltration of sensitive data processed by the application.
Recommendation
- Update LightRAG-HKU to version 1.5.5 or later immediately to patch CVE-2026-85734.
- Implement external rate limiting or a Web Application Firewall (WAF) in front of the LightRAG-HKU service to detect and block high-frequency POST requests to the
/loginendpoint. - Monitor web server logs for anomalous patterns of 401 Unauthorized status codes from single IP addresses directed at the
/loginendpoint.
Immediate actions
Upgrade LightRAG-HKU to version 1.5.5 or later.
Threat Hunt
Search logs for high volumes of 401 status codes targeting /login.
Data: Web server logs
Enrichment needed
- Verify current version of LightRAG-HKU deployed in production. (SOC) Determine scope of vulnerability
Mitigations
Patch LightRAG-HKU to version 1.5.5.
CVE-2026-85734
Detection coverage 1
Detect Excessive Login Attempts to LightRAG-HKU
highDetects potential brute-force attempts against the /login endpoint by counting high volumes of 401 status responses from a single source.
Detection queries are available on the platform. Get full rules →