Microsoft's Open Source Tools Hacked: Password-Stealing Malware Hits AI Developer Repos

Microsoft has pulled dozens of open source GitHub projects after hackers injected password-stealing malware, specifically targeting AI developers. At least 70 repos were disabled, many related to Azure, Claude Code, Gemini's CLI, and VS Code.
Attack Details
According to Cloudsmith and OpenSourceMalware, the malware steals stored credentials when users open compromised tools in AI coding apps. Affected repos include those for Microsoft's cloud services and AI development tooling. OpenSourceMalware identified this as a 're-compromise' of the Durable Task project, which was first breached in mid-May, suggesting the initial cleanup was incomplete or a distinct new attack occurred.
Microsoft's Response
Microsoft spokesperson Ben Hope stated: 'We have temporarily removed some repositories as we investigated potential malicious content. Some repos have been restored after review, while others may remain offline while work continues. As part of our investigation, we notified a small number of customers who may have pulled down content from the affected repositories.'
GitHub displays the message: 'Access to this repository has been disabled by GitHub Staff due to a violation of GitHub's terms of service.'
Implications for Developers
If you have pulled any Microsoft open source tools between mid-May and June 8, 2026, check your credentials stored in your AI development tools. Attackers could have accessed AI developer machines with privileged access to cloud systems and customer data. Consider rotating any passwords or tokens stored in those environments.
This incident is the latest in a series of supply chain attacks targeting popular open source projects. While rare for a large vendor like Microsoft, it underscores that no repository is immune.
📖 Read the full source: HN AI Agents
👀 See Also

jqwik v1.10.0 Sneaks Prompt Injection That Deletes Code When Used by AI Agents
Johannes Link added a hidden instruction to jqwik v1.10.0 that tells AI coding agents to delete all jqwik tests and code, concealed with ANSI escapes. Claude correctly flags it, but human users may not be so lucky.

OpenClaw User Shares Strategy for Balancing Agent Autonomy and Web Security
An OpenClaw user describes their current challenge: balancing agent autonomy with security, particularly regarding web access and prompt injection risks. They propose a solution using 'low trust' and 'high trust' agent segments with a human approval gate.

Anthropic reveals industrial-scale Claude AI data extraction by Chinese labs
Anthropic confirmed Chinese AI labs used over 24,000 fraudulent accounts to scrape 16 million exchanges from Claude, extracting safety guardrails and logic structures for military and surveillance systems.

Claude's Security Review Command Has Limitations for Production Systems
A developer found Claude's security review command helpful for basic validation like MIME types and file size limits, but insufficient for production hardening against sophisticated threats. The solution required a two-week architectural overhaul separating file processing into a restricted worker with limited permissions.