PyPI: anthropic

CVE-2026-34450

Safety vulnerability ID: SFTY-20260401-85495

Safety legacy ID: pyup.io-91354

Affected versions of the anthropic package are vulnerable to Incorrect Default Permissions due to the local filesystem memory tool creating memory files with overly permissive mode 0o666. Both the synchronous and asynchronous memory tool implementations wrote persisted agent state files that were world-readable under a standard umask and world-writable in environments with a permissive umask, such as many Docker base images. A local attacker on a shared host could read sensitive persisted agent state, and in containerised deployments could modify memory files to influence subsequent model behaviour.

Created at: Apr 13, 2026Updated at: Apr 13, 2026

Overview

Claude SDK for Python has Insecure Default File Permissions in Local Filesystem Memory Tool

Advisory

Affected versions of the anthropic package are vulnerable to Incorrect Default Permissions due to the local filesystem memory tool creating memory files with overly permissive mode 0o666. Both the synchronous and asynchronous memory tool implementations wrote persisted agent state files that were world-readable under a standard umask and world-writable in environments with a permissive umask, such as many Docker base images. A local attacker on a shared host could read sensitive persisted agent state, and in containerised deployments could modify memory files to influence subsequent model behaviour.

Affected Package

Affecting anthropic package, versions
>=0.86.0,<0.87.0

Also affects

---

How to Fix

Upgrade
anthropic
to
0.87.0
or higher.

Mitigation and Workarounds

---

Vulnerable Functions

Functions linked to known vulnerabilities.

Vulnerable function data is available for Enterprise customers

Book a call with us to see Safety in action.

Safety

Verified by Safety

Our Cybersecurity Intelligence Team reviewed this vulnerability. We combine public data with our own research to find issues not yet reported to public sources.

Learn more