INSPECTING ARCHIVED INTELLIGENCE (OUTDATED VERSION).
Flaws in Claude Code Put Developers' Machines at Risk
| 2026-02-25 22:02 CRITICAL HIGHExecutive Summary AI-generated
The recent discovery of vulnerabilities in Anthropic's Claude Code AI coding assistant has exposed a significant risk to developers' machines and sensitive information. The flaws, identified as CVE-2025-59536 and CVE-2026-21852, allow attackers to execute commands without user consent and steal API keys from untrusted repositories. This poses a threat of remote code execution and potential data breaches, highlighting the need for immediate action by Anthropic and its users to mitigate these risks.
Technical Mitigations AI-generated
* Use the latest version of Claude Code to ensure protection against security vulnerabilities.
* Regularly review and update project repositories to prevent unauthorized access or malicious modifications.
* Implement additional security features, such as authentication and authorization mechanisms, to harden the coding platform against attacks.
* Monitor project repositories for suspicious activity and report any potential threats to Anthropic's support team.
Intelligence Metadata
Actors / Malware / CVEs / Campaigns
CVE-2025-59536CVE-2025-59536
CVE-2025-59356CVE-2025-59356
CVE-2026-21852CVE-2026-21852
Target & Sectors
Global Scope
Incident Timeline
2025-02-25
Check Point Research identified and reported the flaws to Anthropic.
Click on any entity below to view its context and source!
organisation
Check Point Research
Anthropic fixed the issues after Check Point Research discovered the flaws and reported it to the company last year.
February 25, 2026
Threat actors used untrusted repositories to turn Anthropic's Claude code into an attack vector.
Click on any entity below to view its context and source!
organisation
API
Untrusted repositories turn Claude code into an attack vector
Pierluigi Paganini
February 25, 2026
Flaws in Anthropic’s Claude Code could allow remote code execution and theft of API keys when users open untrusted repositories.
tactic
Remote Code Execution
Untrusted repositories turn Claude code into an attack vector
Pierluigi Paganini
February 25, 2026
Flaws in Anthropic’s Claude Code could allow remote code execution and theft of API keys when users open untrusted repositories.
2026-02-25
Check Point discovered vulnerabilities in Anthropic's Claude Code AI coding assistant, CVE-2025-59536 and CVE-2026-21852, which allowed attackers to execute arbitrary shell commands and exfiltrate API credentials through malicious repository-level configuration files.
Click on any entity below to view its context and source!
organisation
Check Point Research
Check Point Research team found multiple vulnerabilities in Anthropic’s Claude Code AI coding assistant that could lead to remote code execution and API key theft.
organisation
Anthropic’s Claude Code AI
Check Point Research team found multiple vulnerabilities in Anthropic’s Claude Code AI coding assistant that could lead to remote code execution and API key theft.
organisation
CVE-2025
Anthropic has assigned a single identifier,
CVE-2025-59536
, for tracking both flaws.
organisation
CVE-2025-59536
Related:
'God-Like' Attack Machines: AI Agents Ignore Security Policies
The second vulnerability, also tracked as CVE-2025-59536, is associated with Claude Code's
Model Context Protocol
(MCP) setting for connecting the coding platform with external services and tools.
“Critical vulnerabilities, CVE-2025-59536 and CVE-2026-21852, in Anthropic’s Claude Code enabled remote code execution and API key theft through malicious repository-level configuration files, triggered simply by cloning and opening an untrusted project.”
organisation
MCP
Related:
'God-Like' Attack Machines: AI Agents Ignore Security Policies
The second vulnerability, also tracked as CVE-2025-59536, is associated with Claude Code's
Model Context Protocol
(MCP) setting for connecting the coding platform with external services and tools.
The vulnerabilities abuse features such as Hooks, MCP servers, and environment variables to run arbitrary shell commands and exfiltrate Anthropic API credentials when users clone and open untrusted repositories.
organisation
the Model Context Protocol
The risks include silent command execution via abused Hooks, consent bypass in the Model Context Protocol (CVE-2025-59536), and API key exfiltration before trust confirmation (CVE-2026-21852), potentially exposing broader AI-driven workflows.
VIDEO
Anthropic’s API Workspaces feature lets multiple API keys share access to cloud-stored project files.
organisation
API Workspaces
The risks include silent command execution via abused Hooks, consent bypass in the Model Context Protocol (CVE-2025-59536), and API key exfiltration before trust confirmation (CVE-2026-21852), potentially exposing broader AI-driven workflows.
VIDEO
Anthropic’s API Workspaces feature lets multiple API keys share access to cloud-stored project files.
organisation
CVE-2026
“Critical vulnerabilities, CVE-2025-59536 and CVE-2026-21852, in Anthropic’s Claude Code enabled remote code execution and API key theft through malicious repository-level configuration files, triggered simply by cloning and opening an untrusted project.”
infrastructure
2.0.65
The other vulnerability
CVE-2026-21852
, affects Claude Code versions prior to 2.0.65 and allowed API credential theft via malicious project configurations.
organisation
API
The other vulnerability
CVE-2026-21852
, affects Claude Code versions prior to 2.0.65 and allowed API credential theft via malicious project configurations.
organisation
CVE-2025-59356
Configuration Files as Attack Vector
One of the three vulnerabilities that Check Point discovered, CVE-2025-59356, involves a Claude Code feature called Hooks that allows developers to enforce consistent and pre-determined behavior — like code-formatting — at specific points in a project life cycle.
organisation
Attack Vector
Configuration Files as Attack Vector
One of the three vulnerabilities that Check Point discovered, CVE-2025-59356, involves a Claude Code feature called Hooks that allows developers to enforce consistent and pre-determined behavior — like code-formatting — at specific points in a project life cycle.
organisation
Claude Code Put Developers' Machines at Risk
Flaws in Claude Code Put Developers' Machines at Risk.
organisation
React2Shell Exposure
Attackers Use New Tool to Scan for React2Shell Exposure
Two of the vulnerabilities are closely related and involve configuration files in a project repository executing commands without proper user consent.
organisation
GitHub Copilot
Common examples of similar tools include GitHub Copilot, Amazon CodeWhisperer, and OpenAI's Codex.
organisation
Amazon CodeWhisperer
Common examples of similar tools include GitHub Copilot, Amazon CodeWhisperer, and OpenAI's Codex.
organisation
Check Point
Check Point developed an exploit for the vulnerability to show how an adversary could leverage it to gain remote access to a developer's terminal with all the privileges of the developer.
organisation
SecurityAffairs
“As AI integration deepens, security controls must evolve to match the new trust boundaries.”
Follow me on Twitter:
@securityaffairs
and
Facebook
and
Mastodon
Pierluigi Paganini
(
SecurityAffairs
– hacking, Claude)
Tactical Metrics
Metrics
infrastructure
2.0.65
Software Version
Click for context!
The other vulnerability
CVE-2026-21852
, affects Claude Code versions prior to 2.0.65 and allowed API credential theft via malicious project configurations.
Intelligence Sources
Dark Reading
2026-02-25
Security Affairs
2026-02-25
Untrusted repositories turn Claude code into an attack vector
Security Affairs
Unpublish from Social Media?
Are you sure you want to delete this podcast video from all synchronized social networks (YouTube, Facebook, Threads)?
Important:
Due to Meta API restrictions, Instagram Reels cannot be deleted automatically via API by third-party apps.
View Profile to Delete Manually
View Profile to Delete Manually
Tactical Intelligence
Report Intelligence Issue
Podcast Options
Generate
Incident Version History
CURRENT VERSION
Last Updated: 2026-04-27T07:24
Comprehensive Tactical Telemetry
Highly Correlated Entities
17x
organisation
Identified Entity
Check Point Research
entity
3x
vulnerability
Exploited CVE
CVE-2025-59536
cve
2x
timeline
Temporal Reference
2025-02-25
date
2x
tactic
Cyber Operation Type
Exfiltration
tactic
Contextual Telemetry
Context Block
2 METRICS
infrastructure
Software Version
2.0.65
version
tactic
MITRE ATT&CK Technique
T1588.002 - Tool
technique
Click on any entity below to view its context in the main text!
Selective Unpublish
Selecciona las redes de las que quieres eliminar esta publicación. El sistema intentará borrar el post real de la API y limpiará la base de datos para que puedas volver a lanzarlo.
By navigating this website, you accept the use of strictly necessary technical cookies for session security and basic platform functionality. We do not use tracking or advertising cookies.
Read our Privacy Policy.