HighVulnerabilityLLM-specific

CVE-2026-55607: Claude Code worktree handling allows sandbox escape via git directory confusion

Source: NVD(opens in a new tab)
Identifier: CVE-2026-55607
Published: June 29, 2026
Record updated: October 10, 2026

Affected: @anthropic-ai/claude-code >= 2.1.38, < 2.1.163
Fixed in: 2.1.163
Known exploitation: Not listed in the CISA Known Exploited Vulnerabilities catalog at the last check.
EPSS: 0.7%

Summary

Claude Code versions 2.1.38 through 2.1.163 allowed creation of worktrees named ".git" and navigation to worktrees outside the sandbox context, enabling git directory confusion attacks. Through symlink manipulation and git fsmonitor execution during worktree operations, an attacker could overwrite files such as .zshenv in the user's home directory, leading to code execution outside seatbelt sandbox restrictions. Reliable exploitation required the user to clone a malicious repository containing prompt injection content and run Claude Code against it.

Mitigation

Fixed in 2.1.163.

Details

Scores and exploitation

Patch available: Yes
EPSS: 0.7%Estimated probability that this CVE is exploited in the next 30 days, from the FIRST Exploit Prediction Scoring System. Read on October 13, 2026.
Disclosure date: June 29, 2026

Weaknesses and techniques

Weakness (CWE): CWE-22
CWE-59
CWE-78
Attack pattern (CAPEC): CAPEC-126
CAPEC-88
AI threat technique (MITRE ATLAS): AML.T0051

Affected

Packages: @anthropic-ai/claude-code@>= 2.1.38, < 2.1.163 (fixed: 2.1.163)

Classification

Attack type: Prompt injection,
Other
Attack sophistication: Advanced
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-2026-55607

First tracked: June 29, 2026 at 02:07 PM ET, from NVD/CVE Database

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

Classifier confidence: 93% (self-assessed)

Coverage

News, research and other items that mention CVE-2026-55607.

Malicious .git Configs Can Make Claude, Codex, Cursor, and Other AI Agents Run Attacker CodeThe Hacker News · 2026-09-02

Topics

AI agents

Summary

Mitigation

Coverage

Topics

Related items