---
id: "CVE-2026-123851"
page: "https://infrastack.ai/live/CVE-2026-123851"
kind: "vulnerability"
aliases: ["CVE-2026-123851"]
title: "JetBrains Junie GitHub Action: prompt injection protection bypassed via HTML encoding"
severity: "medium"
cvss: 5.4
epss: null
exploited: false
status: "open"
category: "ai-tooling"
products: []
ecosystems: []
tags: ["ai-agents", "ai-tooling", "ci", "github-actions", "junie"]
not_affected_if: ["no-untrusted-input", "version-1.6.4-or-later"]
published_at: "2026-11-06T17:17:24+00:00"
first_seen_at: "2026-11-06T17:43:13+00:00"
updated_at: "2026-11-06T17:45:23+00:00"
sources: [{"source": "nvd", "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-123851", "first_seen_at": "2026-11-06T17:43:13+00:00"}]
exploit_sources: []
---

# JetBrains Junie GitHub Action: prompt injection protection bypassed via HTML encoding

Link for people: https://infrastack.ai/live/CVE-2026-123851 (this .md is for agents)

The JetBrains Junie GitHub Action before 1.6.4 has a prompt injection filter that can be bypassed using HTML-encoded characters. An attacker who can place text in a repository input the agent reads, such as an issue or pull request, may be able to steer the agent's actions.

_Summary, check and fix written by a model from the sources; labelled as such._

## How to check

- Check the version of the Junie GitHub Action referenced in your workflow files (`.github/workflows/`), and confirm whether it is below `1.6.4`.
- Look for workflows that run Junie on issues, pull requests or comments from outside contributors.

## Fix

- JetBrains Junie GitHub Action: move to `1.6.4` or later.

## Timeline

- 2026-11-06T17:43:13+00:00 nvd https://nvd.nist.gov/vuln/detail/CVE-2026-123851

## From the sources

In JetBrains Junie GitHub Action before 1.6.4 prompt injection protection could be bypassed via HTML-encoded characters

---

_Free to read, with fair use of about 60 pages a minute. For higher limits or API access: sales@infrastack.ai_