CVE-2026-122134 · AI agents & MCP
NO FIX YET · SEE WHAT TO DO
Mistral Vibe, a coding agent, runs git status at startup, which executes malicious Git clean or process filter drivers in a repository. An attacker who gets a user to open a crafted repository runs commands with that user's privileges before any trust prompt appears.
first public Nov 4, 2026 · 1 source
Check it with your coding agent
It checks your repository and tells you if you’re affected, with the fix.
Developers who open untrusted or third-party Git repositories with Mistral Vibe before version 2.25.7 are exposed, because the git status call at startup triggers the filter drivers. Teams that only open their own trusted repositories are at lower risk. The sources do not say whether any non-default setting is needed.
Not affected if trusted-repos-only, version-2.25.7-or-later
Who cares ai-coding-agent, git, mistral-vibe
2.25.7 are listed as affected.2.25.7 or later.Mistral Vibe before 2.25.7 contains a remote code execution vulnerability that allows attackers to run commands by configuring malicious Git clean or process filter drivers in a repository. When a user opens a crafted repository, ProjectContextProvider._fetch_git_status runs git status at startup, invoking the filters and executing commands with user privileges before any trust prompt.
Are you affected?
Tell InfraStack what you run. It checks this one, and every next one, against your stack.
Ask from here. InfraStack checks the affected versions against your stack and answers in your inbox.
InfraStack checks the affected versions against your stack and answers in Slack, with the evidence and the fix.
NVD data is provided by NIST and is public domain. Posts quoted belong to their authors.