← Vulnerability feed

Vulnerability record · CVE-2026-44246 · published 12 May 2026

CVE-2026-44246: Dkfz nnu-net injection vulnerability

DDkfz · Nnu Net

nnU-Net is a semantic segmentation framework that automatically adapts its pipeline to a dataset. Prior to 2.4.1, the nnU-Net Issue Triage workflow in .github/workflows/issue-triage.yml is vulnerable to Agentic Workflow Injection. The workflow sets allowed_non_write_users: ${{ github.event.issue.user.login }}, which means any logged-in GitHub user who opens an issue can reach this agentic workflow with attacker-controlled content. Untrusted issue title and body content are embedded directly into the prompt of anthropics/claude-code-action, and the workflow then runs a command-capable Claude agent with permission to comment on and relabel the current issue via gh. Because this workflow is triggered automatically on issues.opened, an external attacker can submit a crafted issue that steers the agent beyond its intended issue-triage purpose and influences authenticated issue actions. This vulnerability is fixed in 2.4.1.

7.2 CVSS 3.1 High EPSS 0.29% · top 80.6% CWE-74 · Injection
7.2CVSS 3.1 base score
0.29%EPSS exploitation probability, 30 days
NoNot in CISA KEV
1Affected product versions listed by NVD
2References, 2 tagged exploit
17 Jun 2026Last modified by NVD

Description

nnU-Net is a semantic segmentation framework that automatically adapts its pipeline to a dataset. Prior to 2.4.1, the nnU-Net Issue Triage workflow in .github/workflows/issue-triage.yml is vulnerable to Agentic Workflow Injection. The workflow sets allowed_non_write_users: ${{ github.event.issue.user.login }}, which means any logged-in GitHub user who opens an issue can reach this agentic workflow with attacker-controlled content. Untrusted issue title and body content are embedded directly into the prompt of anthropics/claude-code-action, and the workflow then runs a command-capable Claude agent with permission to comment on and relabel the current issue via gh. Because this workflow is triggered automatically on issues.opened, an external attacker can submit a crafted issue that steers the agent beyond its intended issue-triage purpose and influences authenticated issue actions. This vulnerability is fixed in 2.4.1.

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

Affected products

1 vulnerable configurations from NVD's CPE data, grouped by vendor and product.

References

Track CVE-2026-44246 inside VULONE

Watch it alongside the ransomware crews, C2 infrastructure and forum chatter that reference it, query it through the API and pull it into your SIEM over TAXII.

Start free Open in platform

Related vulnerabilities

Same products first, then exploited flaws of the same weakness class.

10.0CVE-2025-20337Cisco ISE API input validation flaw allows unauthenticated root RCECisco ISE and ISE-PIC fail to properly validate user-supplied input in a specific API, letting an unauthenticated remote attacker execute arbitrary c…KEVEPSS 68%analysed10.0CVE-2025-20281Cisco ISE API unauthenticated remote code executionCisco ISE and ISE-PIC fail to properly validate user-supplied input in a specific API, allowing a crafted request to reach the underlying operating s…KEVEPSS 98%analysed7.2CVE-2022-43769Hitachi Vantara Pentaho BA Server Spring Template InjectionHitachi Vantara Pentaho Business Analytics Server before 9.4.0.1 and 9.3.0.2, including 8.3.x, allows certain web services to set property values con…KEVEPSS 98%analysed9.8CVE-2023-22527Atlassian Confluence Data Center and Server template injection RCEOlder versions of Confluence Data Center and Server contain a template injection flaw (CWE-74) that lets an unauthenticated attacker execute code on …KEVEPSS 100%analysed9.8CVE-2022-35914GLPI htmlawed Test Script PHP Code InjectionThe htmLawed test script shipped inside GLPI up to 10.0.2 allows PHP code injection through the htmlawed module. Because the vulnerable file is reach…KEVEPSS 100%analysed9.8CVE-2022-46169Cacti remote_agent.php auth bypass leads to OS command injectionCacti's remote_agent.php trusts attacker-controlled HTTP headers when resolving the client IP, letting an unauthenticated attacker spoof the poller h…KEVEPSS 100%analysed7.8CVE-2018-19322GIGABYTE driver IO port access allows privilege escalationThe GPCIDrv and GDrv low-level drivers shipped with GIGABYTE APP Center, AORUS GRAPHICS ENGINE, XTREME GAMING ENGINE and OC GURU II expose IO port re…KEVEPSS 1.8%analysed7.5CVE-2022-27924Zimbra Collaboration memcache command injection via unauthenticated requestZimbra Collaboration Suite 8.8.15 and 9.0 allows an unauthenticated attacker to inject arbitrary memcache commands into a targeted instance. The inje…KEVEPSS 85%analysed

Source: NIST National Vulnerability Database (record CVE-2026-44246), CISA KEV, FIRST EPSS (scores of 2026-09-26). This page is refreshed as NVD updates the record.