Skip to main content

Claude Sdk For Python CVE-2026-34452

| EUVDEUVD-2026-17681 MEDIUM
Improper Link Resolution Before File Access (CWE-59)
2026-03-31 security-advisories@github.com GHSA-w828-4qhx-vxx3
5.8
CVSS 4.0 · Vendor: github
View

Vulnerability Assessment

Weakness
CWE
OWASP Top 10:2025
A01:2025 - Broken Access Control
ASVS 5.0
V5 - File Handling Symlink resolution safety
Impact

CVSS score of 5.8 reflects moderate risk (local attack vector, high privilege requirement at layer L, high confidentiality impact). However, EPSS and exploitation probability are not provided in this dataset. The vulnerability requires local filesystem access and authenticated SDK usage, substantially limiting threat surface compared to remote exploits. Real-world impact depends on deployment context: in multi-tenant or shared-host scenarios where untrusted users can write to the memory directory, this is a meaningful information disclosure risk. The TOCTOU nature means exploitation requires precise timing or repeated attempts, increasing attack complexity (AC:H in CVSS). No public exploit code or active exploitation signals are evident, but the straightforward nature of symlink manipulation suggests proof-of-concept development is trivial. Remediation is simple (upgrade to 0.87.0), making this a moderate-priority patch rather than critical-prioritize if hosting multi-user Python SDK applications on shared infrastructure.

Exploit ScenarioDerived

An attacker with local shell access and write permissions to an application's memory directory monitors the directory for in-flight file operations. When the async memory tool validates a path like /memory/data.txt, the attacker quickly replaces a symlink pointing to /memory/data.txt with one targeting /etc/passwd or another sensitive file. The validation passes (the symlink initially pointed inside the sandbox), but the file read returns contents from /etc/passwd, disclosing sensitive configuration data. Alternatively, a write operation could corrupt or modify files outside the sandbox. Exploitation requires precise timing (AC:H) but is feasible in environments with predictable workload patterns or via polling techniques.

This Month Plan within 30 days. Why: MEDIUM severity (CVSS 4.0-6.9).
Patch available 0.0% exploitation probability (EPSS)

Get an e-mail when this CVE is exploited, patched or re-scored. Free account.

Alert me

AnalysisAIv1 – Mar 31, 2026

Symlink race condition in Anthropic Python SDK async filesystem memory tool (versions 0.86.0-0.86.x) allows local authenticated attackers to escape sandbox restrictions and read or write arbitrary files outside the designated memory directory. The vulnerability exploits a time-of-check-time-of-use (TOCTOU) flaw where path validation occurs before symlink resolution, enabling an attacker with memory directory write access to redirect file operations via symlink manipulation. The synchronous implementation is unaffected. Vendor-released patch: version 0.87.0.

RemediationAI

Upgrade Anthropic Python SDK to version 0.87.0 or later immediately. The patch resolves the vulnerability by properly resolving symlinks before performing file operations, eliminating the TOCTOU window. Users unable to upgrade immediately should restrict write permissions on the memory directory to the application process only, preventing untrusted local users from creating or modifying symlinks. The fix is included in the released version 0.87.0 available on PyPI; apply via pip install --upgrade anthropic>=0.87.0. Reference commit 6599043eee6e86dce16953fcd1fd828052052be6 in the upstream repository for implementation details.

Affected VersionsENISA

anthropic-sdk-python 0.86.0, < 0.87.0

Affected Packages

pypi anthropic fix: 0.87.0
euvd anthropic_sdk_python_0_86_0 fix: 0.87.0

Source: OSV (Open Source Vulnerabilities)

Affects your stack

Affects 2 packages in pypi, euvd

Checking this CVE against your own stacks and servers is part of the Basic plan and up.

Technical ContextAI · show details

The Anthropic Python SDK implements local filesystem memory tools to provide sandboxed file access for Claude API applications. The async variant validates that model-supplied file paths resolve within a designated memory directory sandbox, but critically passes the unresolved path to subsequent file operations rather than the canonicalized result. This creates a classic TOCTOU race condition (CWE-59) where a local attacker with write permissions to the memory directory can modify symlinks between the validation check and actual file I/O. An attacker exploiting this can craft symlinks that pass validation (pointing inside the sandbox) but are retargeted to arbitrary locations before the file operation executes, bypassing all sandbox enforcement. The synchronous implementation avoids this issue by likely using resolved paths for operations.

Know when Claude Sdk For Python gets its next vulnerability

  • Watch products and get an email when something new is published
  • Check this CVE against your own stack and servers
  • Exploit and PoC references, threat intelligence from 22 sources

No credit card · 7-day full trial

Severity by source

Vendor (github) PRIMARY 5.8 MEDIUM
CVSS:4.0/AV:L/AC:H/AT:N/PR:L/UI:N/VC:H/VI:L/VA:N/SC:N/SI:N/SA:N

Primary rating from Vendor (github) · only source for this CVE.

CVSS VectorVendor: github

Attack Vector
Local
Attack Complexity
High
Attack Requirements
None
Privileges Required
Low
User Interaction
None
Vulnerable system: Confidentiality
High
Vulnerable system: Integrity
Low
Vulnerable system: Availability
None
Subsequent system: Confidentiality
None
Subsequent system: Integrity
None
Subsequent system: Availability
None

Lifecycle Timeline

2
Patch released
Apr 02, 2026 - 02:30 nvd
Patch available
CVE Published
Mar 31, 2026 - 22:16 nvd
MEDIUM 5.8

DescriptionCVE.org

The Claude SDK for Python provides access to the Claude API from Python applications. From version 0.86.0 to before version 0.87.0, the async local filesystem memory tool in the Anthropic Python SDK validated that model-supplied paths resolved inside the sandboxed memory directory, but then returned the unresolved path for subsequent file operations. A local attacker able to write to the memory directory could retarget a symlink between validation and use, causing reads or writes to escape the sandbox. The synchronous memory tool implementation was not affected. This issue has been patched in version 0.87.0.

Threat Intelligence Overview

EPSS Score
0.0%
Top 97% exploitable
Threat Intel
1
1 source
Data Quality
6 sources
Confidence: High
EUVD NVD CVE.org EPSS Vulnrichment GHSA GitHub VulDB

How this decision was reached

The engine reads each signal below and applies a fixed list of rules from top to bottom. The first rule that matches sets the recommended action – like a statement, every line is shown.

SignalValue
Severity MEDIUM
CISA KEV No
Public POC No
EPSS 0.0%
Patch available Yes
Active threat No
Deciding rule
MEDIUM severity (CVSS 4.0-6.9)
What every action level means
Emergency Immediate action required. Criteria: CRITICAL + in CISA KEV (actively exploited) without patch, OR CRITICAL + EPSS > 30% without patch, OR CRITICAL + active threat signals without patch.
Act Now Prioritize within 24-48h. Criteria: in KEV with patch available, OR CRITICAL + public POC, OR HIGH + EPSS > 10%, OR CRITICAL + active threat, OR any CRITICAL severity.
This Week Schedule remediation this week. Criteria: HIGH severity (CVSS 7.0-8.9). Patch within normal change window.
This Month Plan within 30 days. Criteria: MEDIUM severity (CVSS 4.0-6.9), lower exploitation probability. Include in regular patch cycle.
Monitor Track and reassess. Criteria: LOW severity (CVSS 0.1-3.9), minimal exploitation risk. No immediate action needed.

SSVC AssessmentENISA

SSVC Decision· derived
Defer
Exploitation
None
System Exposure· derived
Small
Automatable
No
Human Impact· derived
Medium
Technical Impact
Partial

Priority Score

29
Low Medium High Critical
KEV: 0
EPSS: 0
CVSS: +29
PoC: 0
Weaponized: 0
Threat intel: 0
Gov advisory: 0

Attack Technique

Information disclosure occurs when an application unintentionally exposes sensitive data that aids attackers in reconnaissance or directly compromises security.

This site uses cookies essential for authentication and security. No tracking or analytics cookies are used. Privacy Policy