CVE-2026-33068
Received
Received - Intake
Permission Bypass in Claude Code via Malicious Settings File
Publication date: 2026-03-20
Last updated on: 2026-03-24
Assigner: GitHub, Inc.
Description
Description
Claude Code is an agentic coding tool. Versions prior to 2.1.53 resolved the permission mode from settings files, including the repo-controlled .claude/settings.json, before determining whether to display the workspace trust confirmation dialog. A malicious repository could set permissions.defaultMode to bypassPermissions in its committed .claude/settings.json, causing the trust dialog to be silently skipped on first open. This allowed a user to be placed into a permissive mode without seeing the trust confirmation prompt, making it easier for an attacker-controlled repository to gain tool execution without explicit user consent. This issue has been patched in version 2.1.53.
CVSS Scores
EPSS Scores
| Probability: | |
| Percentile: |
Meta Information
Affected Vendors & Products
| Vendor | Product | Version / Range |
|---|---|---|
| anthropic | claude_code | to 2.1.53 (exc) |
Helpful Resources
Exploitability
| CWE ID | Description |
|---|---|
| CWE-807 | The product uses a protection mechanism that relies on the existence or values of an input, but the input can be modified by an untrusted actor in a way that bypasses the protection mechanism. |