HighVulnerabilityLLM-specific

CVE-2026-44246: nnU-Net issue triage workflow prompt injection via crafted GitHub issue

Source: NVD(opens in a new tab)
Identifier: CVE-2026-44246
Published: May 12, 2026
Record updated: October 9, 2026

Known exploitation: Not listed in the CISA Known Exploited Vulnerabilities catalog at the last check.
EPSS: 0.3%

Summary

CVE-2026-44246 affects nnU-Net, a semantic segmentation framework, before version 2.4.1. Its Issue Triage workflow in .github/workflows/issue-triage.yml sets allowed_non_write_users to the issue author's login, so any logged-in GitHub user who opens an issue can reach the workflow. Untrusted issue title and body text is embedded directly into the prompt of anthropics/claude-code-action, which runs a command-capable Claude agent that can comment on and relabel the issue via gh, and the workflow triggers automatically on issues.opened. An external attacker can therefore submit a crafted issue that steers the agent beyond issue triage and influences authenticated issue actions.

Mitigation

Fixed in 2.4.1.

Details

Scores and exploitation

CVSS score: 7.2(high)
CVSS vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:L/I:L/A:N
Attack vector: network
Attack complexity: low
Privileges required: none
User interaction: none
EPSS: 0.3%Estimated probability that this CVE is exploited in the next 30 days, from the FIRST Exploit Prediction Scoring System. Read on October 12, 2026.
Disclosure date: May 12, 2026

Weaknesses and techniques

AI threat technique (MITRE ATLAS): AML.T0051

Classification

Attack type: Prompt injection
Attack sophistication: Trivial
AI component targeted: Agent
Impact: Integrity,
Confidentiality
Labels: security

How this record was produced

Summaries, mitigations and labels on this page are generated automatically from the source. The linked source is authoritative. The source gives this advisory a description and no title, so the title above is a headline generated from the source text.

Original source: https://nvd.nist.gov/vuln/detail/CVE-2026-44246

First tracked: May 12, 2026 at 08:09 PM ET, from NVD/CVE Database

Classifier: claude-haiku-5-5, prompt v4

Classifier confidence: 90% (self-assessed)

Summary

Mitigation

Related items