What is the CVSS score of EUVD-2026-33430?

EUVD-2026-33430 has a CVSS 3.1 base score of 7.7 (High). CVSS vector: CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:N/A:N. EPSS exploitation probability: 0.0%.

Which versions of FastGPT are affected by EUVD-2026-33430?

Labring FastGPT versions prior to 4.15.0-beta1 contain a critical network isolation bypass that allows authenticated users to probe and access internal services, creating a backdoor for lateral…. A patch is available.

How to fix or mitigate EUVD-2026-33430?

Within 24 hours: Identify all FastGPT deployments in your environment and audit dataset preview endpoint logs for suspicious internal IP access patterns; immediately restrict dataset preview access to users with documented business need. Within 7 days: Implement network firewall rules isolating FastGPT instances and preventing them from reaching internal services; deploy WAF rules blocking GET requests to internal IP ranges from the preview endpoint. Within 30 days: Establish continuous monitoring for dataset preview endpoint activity; evaluate architectural redesign to further isolate FastGPT from internal networks pending vendor patch availability.

FastGPT EUVD-2026-33430

| CVE-2026-44285 HIGH

Server-Side Request Forgery (SSRF) (CWE-918)

2026-05-29 GitHub_M

SSRF Fastgpt

7.7

CVSS 3.1 · Vendor: GitHub_M

Severity by source

Vendor (GitHub_M) PRIMARY

7.7 HIGH

AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:N/A:N

Primary rating from Vendor (GitHub_M) · only source for this CVE.

CVSS VectorVendor: GitHub_M

CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:N/A:N

Attack Vector

Network

Attack Complexity

Low

Privileges Required

Low

User Interaction

None

Scope

Changed

Confidentiality

High

Integrity

None

Availability

None

Lifecycle Timeline

Patch available

May 29, 2026 - 21:02 EUVD

Analysis Generated

May 29, 2026 - 20:30 vuln.today

DescriptionCVE.org

FastGPT is an AI Agent building platform. Prior to 4.15.0-beta1, a Server-Side Request Forgery (SSRF) vulnerability allows an authenticated attacker to bypass the global isInternalAddress network protection and make arbitrary HTTP GET requests to internal network services. This is achieved by exploiting an incomplete fix in the dataset preview endpoint /api/core/dataset/file/getPreviewChunks when utilizing the externalFile data import type. This vulnerability is fixed in 4.15.0-beta1.

AnalysisAI

Server-Side Request Forgery in Labring FastGPT prior to 4.15.0-beta1 lets an authenticated attacker bypass the platform's isInternalAddress network protection and pivot HTTP GET probes into internal services via the dataset preview endpoint. The flaw stems from an incomplete prior fix in the externalFile data import path, scoped-changed impact (S:C) elevates risk to adjacent systems, and no public exploit identified at time of analysis.

Technical ContextAI

FastGPT is an open-source AI agent / RAG building platform (cpe:2.3:a:labring:fastgpt) that ingests datasets from multiple source types including 'externalFile'. The vulnerability is a CWE-918 (Server-Side Request Forgery) flaw in the /api/core/dataset/file/getPreviewChunks endpoint: a previously deployed isInternalAddress allow/deny list was intended to block requests to RFC1918/loopback/link-local destinations, but the externalFile import path did not consistently apply that check, allowing a server-issued HTTP GET to traverse to internal-only services. SSRF in an agent platform is particularly impactful because backend hosts frequently sit alongside cloud metadata endpoints, internal admin APIs, vector DBs, and orchestration services.

RemediationAI

Vendor-released patch: upgrade to FastGPT 4.15.0-beta1 or later as documented in GHSA-c65v-7vx6-f8m3 (https://github.com/labring/FastGPT/security/advisories/GHSA-c65v-7vx6-f8m3). If an immediate upgrade is not possible, restrict access to the /api/core/dataset/file/getPreviewChunks endpoint at the reverse proxy and disable or block the externalFile dataset import type for untrusted tenants, accepting that legitimate URL-based dataset previews will stop working. Additionally place the FastGPT backend in a network segment that egress-filters traffic to RFC1918, 169.254.169.254 (cloud metadata), and other internal management planes, and where supported enforce IMDSv2 on the underlying cloud instance to neutralize credential-theft scenarios; note this hardens the blast radius but does not remediate the SSRF itself.