Claude Agent SDK: Argument Injection via resume Option Allows Arbitrary Command Execution
- Product
- claude-agent-sdk
- CVE
- CVE-2026-96620
- Affected
- claude-agent-sdk < 0.2.121
- Fixed in
- 0.2.121
Last verified against sources: October 26, 2026
Am I affected?
claude-agent-sdk: affected claude-agent-sdk < 0.2.121. Fixed in 0.2.121.
- Check your version: run
pip show claude-agent-sdk. Versions < 0.2.121 of claude-agent-sdk are affected. [1]
What to do now
- Upgrade to the fixed version: run
pip install --upgrade "claude-agent-sdk>=0.2.121"(first patched version 0.2.121). [1] - Upgrade to the latest version of the claude-agent-sdk package. [1]
- The first patched version is 0.2.121. [1]
General practice, not from the source: if the affected software could read tokens or credentials, rotate them after upgrading; pin versions and review changes before updating tools your agents run.
What happened
Impact
What was done
Sources
- GHSA-h4mw-j7qp-8mwm — GitHub Advisory Database
Every statement above was checked against these sources before publishing. Spotted an error? Tell us and we will correct it with a dated note.
Share advisory brief with your team
Copy a plain-text summary for Slack, GitHub issues, or team security channels.
[CVE-2026-96620] Claude Agent SDK: Argument Injection via resume Option Allows Arbitrary Command Execution
Product: claude-agent-sdk | Affected: claude-agent-sdk < 0.2.121 | Fixed in: 0.2.121
Summary: The Claude Agent SDK for Python passed the `resume` option to the Claude CLI without checking the value.
Source & remediation on AgentArmorHQ: https://agentarmorhq.com/incidents/2026/10/cve-2026-96620-claude-agent-sdkCorrections
Report an error or request a correction. We review requests within 5 business days and add a dated correction note.
Discussions for CVE-2026-96620
Ask questions or share configuration tips. Governed by our Community Rules & Terms.
No posts match this filter yet.
Be the first to ask a question or share a working tip.