---
id: "CVE-2026-116843"
kind: "vulnerability"
aliases: ["CVE-2026-116843"]
title: "Mistral Vibe: untrusted git repo runs attacker commands on worktree start"
severity: "high"
cvss: 8.8
epss: null
exploited: false
status: "open"
category: "ai-tooling"
products: []
ecosystems: []
tags: ["ai-tooling", "coding-agent", "developer-workstations", "git", "mistral-vibe"]
not_affected_if: ["fixed-version", "no-worktree-flag", "trusted-repos-only"]
published_at: "2026-10-27T01:19:07+00:00"
first_seen_at: "2026-10-27T01:34:33+00:00"
updated_at: "2026-10-27T01:36:13+00:00"
sources: [{"source": "nvd", "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-116843", "first_seen_at": "2026-10-27T01:34:33+00:00"}]
---

# Mistral Vibe: untrusted git repo runs attacker commands on worktree start

Mistral Vibe before 3.0.3 reads git filter driver definitions from an untrusted repository's .git/config when started with --worktree. The attacker's smudge, clean or process commands run with the user's privileges before the trusted-folder check.

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

## How to check

- Check the installed Mistral Vibe version with `vibe --version` (exact command unknown; confirm in the package manifest)
- Look for Mistral Vibe runs using the `--worktree` flag in shell history or launch scripts
- Inspect untrusted repositories' `.git/config` for `filter.*` entries with smudge, clean or process commands

## Fix

- Upgrade Mistral Vibe to `3.0.3` or later
- Until upgraded, avoid running Mistral Vibe with `--worktree` on repositories you did not create

## Timeline

- 2026-10-27T01:34:33+00:00 nvd https://nvd.nist.gov/vuln/detail/CVE-2026-116843

## From the sources

Mistral Vibe before 3.0.3 contains a remote code execution vulnerability that allows attackers to run commands by defining git filter drivers in an untrusted repository's .git/config. When started with --worktree, git worktree add applies .gitattributes and runs filter smudge, clean, or process commands before the trusted-folder check, executing attacker commands with user privileges.

---

_Free to read at a human pace. An API with keys and an MCP server are in preparation; early access: hello@infrastack.ai_