External risk intelligence

LightRAG Login Endpoint Rate Limit Bypass Allows Credential Theft

CVE advisorySeverity: CRITICAL (CVSS 9.1)

CVE-2026-85734

The vulnerability resides in a /login endpoint of a server application designed for retrieval-augmented generation. Such services are commonly deployed as web-facing APIs or management interfaces to facilitate remote access for document retrieval and administrative tasks, making them accessible via the network in typical deployment scenarios.

Halo Surface Signal: 4 out of 5 — likely to be public-facing.

External exposure likelihood

Horizon Alert

Summary of the vulnerability and why it matters

A critical vulnerability has been identified in the LightRAG login process, affecting how it handles authentication attempts. This flaw could allow an attacker to repeatedly guess credentials without detection, potentially leading to unauthorized access to sensitive information and administrative controls. The primary concern is to confirm if your systems utilize this technology and if they are exposed.

  • Unprotected logins allow attackers to guess passwords.
  • Potential access to sensitive documents and admin controls.
  • Confirm relevance and exposure to sensitive data.

Attack Path

How an attacker could exploit the issue

A network attacker can repeatedly guess passwords against the login endpoint without restrictions, eventually discovering valid credentials. This unauthorized access then allows them to view documents, interact with the knowledge graph, and perform administrative actions.

  • No authentication needed to start.
  • Repeatedly guess passwords on login.
  • Gain access to sensitive data.

Live Threat

Current exploitation, exposure, and threat context

An attacker could repeatedly guess passwords for the login endpoint until they find a valid one. This could grant them authenticated access to sensitive information and administrative functions.

  • Documents and knowledge graph data at risk.
  • Brute-force password guessing.
  • Unauthorized access to sensitive information.

Operational Fix

Recommended remediation, mitigation, and detection steps

The LightRAG application owner is responsible for this vulnerability. The first practical step is to identify all instances of LightRAG, determine their network exposure and business criticality, and confirm the accountable owner. Once identified, a risk-based remediation plan can be developed, potentially involving coordination with the vendor for updates.

  • Identify LightRAG instances and owners.
  • Verify network exposure and criticality.
  • Plan remediation based on risk.

Supplementary metadata

Validate whether this threat affects your internet-facing exposure.

Halo Threat Intelligence helps prioritize remediation with Halo Surface Signal and H/A/L/O context. Start exposure validation with a free external attack surface trial.

Frequently asked questions

What is LightRAG?

LightRAG is an open-source software framework used to build Retrieval-Augmented Generation (RAG) applications. It helps developers create systems that can search, retrieve, and reason over large collections of documents by building and querying a knowledge graph, which enhances the accuracy and context of AI-generated responses.

What does CWE-307 mean for CVE-2026-85734?

CWE-307 refers to Improper Restriction of Excessive Authentication Attempts. In plain terms, this means the software lacks defenses like account lockouts or rate limiting. Because of this, CVE-2026-85734 allows an attacker to automate password guessing at high speeds without the system stopping or slowing them down, making it much easier to compromise a valid account.

How do attackers trigger this vulnerability?

An attacker triggers this by sending continuous, automated password guesses to the POST /login endpoint. The vulnerability is specifically caused by the absence of speed limits or failed-attempt counters. It is important to note that successfully logging in with correct credentials through legitimate, non-automated use does not trigger the bug; the risk arises specifically from high-volume, repeated login attempts.

Is my instance of LightRAG at risk?

According to Halo Surface Signal, this vulnerability is most relevant if your LightRAG instance is deployed as a web-facing API or an accessible management interface. Since these services are often intentionally put on the network to allow remote document retrieval, they are likely to be reachable by external actors. Systems restricted to internal, private networks face a different risk profile.

How should I respond to this LightRAG security flaw?

Your first step is to locate all LightRAG deployments in your environment and determine who owns them. Assess whether these instances are reachable over the network and what kind of data they handle. Once you have an inventory, prioritize updates for any systems running versions prior to 1.5.5, which is where this specific authentication weakness is resolved.

References