<feed xmlns="http://www.w3.org/2005/Atom"> <id>https://danusminimus.github.io/</id><title>danusminimus</title><subtitle>A blog dedicated to reverse engineering, malware analysis, and security research.</subtitle> <updated>2026-08-23T05:16:03-04:00</updated> <author> <name>Dan</name> <uri>https://danusminimus.github.io/</uri> </author><link rel="self" type="application/atom+xml" href="https://danusminimus.github.io/feed.xml"/><link rel="alternate" type="text/html" hreflang="en" href="https://danusminimus.github.io/"/> <generator uri="https://jekyllrb.com/" version="4.4.1">Jekyll</generator> <rights> © 2026 Dan </rights> <icon>/assets/img/favicons/favicon.ico</icon> <logo>/assets/img/favicons/favicon-96x96.png</logo> <entry><title>A WIF Of Fresh Access: How a GitHub Issue on Gemini-CLI Led to GCP Project Compromise</title><link href="https://danusminimus.github.io/posts/A-WIF-Of-Fresh-Access-How-A-GitHub-Issue-On-Gemini-CLI-Led-To-GCP-Project-Compromise/" rel="alternate" type="text/html" title="A WIF Of Fresh Access: How a GitHub Issue on Gemini-CLI Led to GCP Project Compromise" /><published>2026-08-18T00:00:00-04:00</published> <updated>2026-08-18T00:00:00-04:00</updated> <id>https://danusminimus.github.io/posts/A-WIF-Of-Fresh-Access-How-A-GitHub-Issue-On-Gemini-CLI-Led-To-GCP-Project-Compromise/</id> <content type="text/html" src="https://danusminimus.github.io/posts/A-WIF-Of-Fresh-Access-How-A-GitHub-Issue-On-Gemini-CLI-Led-To-GCP-Project-Compromise/" /> <author> <name>Dan</name> </author> <category term="AI Security Research" /> <category term="Supply Chain" /> <summary>Executive Summary As traditional applications continue to embed agentic components into their workflows, the Pillar Security team continues to uncover security issues that emerge from the way these systems interact with code, infrastructure, credentials, and untrusted input. In this research, we examine a vulnerability affecting the google/gemini-cli ecosystem. Gemini CLI is Google’s open-s...</summary> </entry> <entry><title>I'll Just Call You: Agent-to-Agent Privilege Boundary Failures in CI/CD on Google's ADK Repository</title><link href="https://danusminimus.github.io/posts/Ill-Just-Call-You-Agent-To-Agent-Privilege-Boundary-Failures-In-CI-CD-On-Googles-ADK-Repository/" rel="alternate" type="text/html" title="I&amp;apos;ll Just Call You: Agent-to-Agent Privilege Boundary Failures in CI/CD on Google&amp;apos;s ADK Repository" /><published>2026-08-03T00:00:00-04:00</published> <updated>2026-08-03T00:00:00-04:00</updated> <id>https://danusminimus.github.io/posts/Ill-Just-Call-You-Agent-To-Agent-Privilege-Boundary-Failures-In-CI-CD-On-Googles-ADK-Repository/</id> <content type="text/html" src="https://danusminimus.github.io/posts/Ill-Just-Call-You-Agent-To-Agent-Privilege-Boundary-Failures-In-CI-CD-On-Googles-ADK-Repository/" /> <author> <name>Dan</name> </author> <category term="AI Security Research" /> <category term="Supply Chain" /> <summary>Executive Summary Pillar Security researchers have identified the first practical, real-world case of agent-to-agent exploitation in a multi-agent system in a real production environment, a class of attack not seen in real production systems until now. A case where one AI agent can be used to attack another, turning a benign automation into a path that ends in a potential software supply chain...</summary> </entry> <entry><title>My Agentic Trust Issues: From Prompt Injection to Supply-Chain Compromise on gemini-cli</title><link href="https://danusminimus.github.io/posts/My-Agentic-Trust-Issues-From-Prompt-Injection-To-Supply-Chain-Compromise-On-gemini-cli/" rel="alternate" type="text/html" title="My Agentic Trust Issues: From Prompt Injection to Supply-Chain Compromise on gemini-cli" /><published>2026-07-08T00:00:00-04:00</published> <updated>2026-07-09T01:22:07-04:00</updated> <id>https://danusminimus.github.io/posts/My-Agentic-Trust-Issues-From-Prompt-Injection-To-Supply-Chain-Compromise-On-gemini-cli/</id> <content type="text/html" src="https://danusminimus.github.io/posts/My-Agentic-Trust-Issues-From-Prompt-Injection-To-Supply-Chain-Compromise-On-gemini-cli/" /> <author> <name>Dan</name> </author> <category term="AI Security Research" /> <category term="Supply Chain" /> <summary>Executive Summary I identified a CVSS 10 critical vulnerability (dubbed TrustIssues) in Google’s AI powered GitHub workflows that allowed any external attacker, with nothing more than a public GitHub issue, to a full supply chain compromise of the gemini-cli repository, Google’s AI coding agent with 101,000+ stars. The critical severity rating reflects a specific bypass I identified inside Ge...</summary> </entry> <entry><title>Prompt Injection leads to RCE and Sandbox Escape in Antigravity</title><link href="https://danusminimus.github.io/posts/Prompt-Injection-Leads-To-RCE-And-Sandbox-Escape-In-Antigravity/" rel="alternate" type="text/html" title="Prompt Injection leads to RCE and Sandbox Escape in Antigravity" /><published>2026-04-26T00:00:00-04:00</published> <updated>2026-07-09T01:22:07-04:00</updated> <id>https://danusminimus.github.io/posts/Prompt-Injection-Leads-To-RCE-And-Sandbox-Escape-In-Antigravity/</id> <content type="text/html" src="https://danusminimus.github.io/posts/Prompt-Injection-Leads-To-RCE-And-Sandbox-Escape-In-Antigravity/" /> <author> <name>Dan</name> </author> <category term="AI Security Research" /> <category term="Agentic IDEs" /> <summary>Executive Summary I discovered a vulnerability in Antigravity, Google’s agentic IDE. This technique exploits insufficient input sanitization of the find_by_name tool’s Pattern parameter, allowing attackers to inject command-line flags into the underlying fd utility, converting a file search operation into arbitrary code execution. Critically, this vulnerability bypasses Antigravity’s Secure M...</summary> </entry> <entry><title>Cursor Allowlist Bypass (CVE-2026-22708)</title><link href="https://danusminimus.github.io/posts/The-Agent-Security-Paradox-When-Trusted-Commands-In-Cursor-Become-Attack-Vectors/" rel="alternate" type="text/html" title="Cursor Allowlist Bypass (CVE-2026-22708)" /><published>2026-01-14T00:00:00-05:00</published> <updated>2026-07-09T01:22:07-04:00</updated> <id>https://danusminimus.github.io/posts/The-Agent-Security-Paradox-When-Trusted-Commands-In-Cursor-Become-Attack-Vectors/</id> <content type="text/html" src="https://danusminimus.github.io/posts/The-Agent-Security-Paradox-When-Trusted-Commands-In-Cursor-Become-Attack-Vectors/" /> <author> <name>Dan</name> </author> <category term="AI Security Research" /> <category term="Agentic IDEs" /> <summary>Executive Summary I discovered a critical vulnerability in Cursor (CVE-2026-22708) that exploits how agentic IDEs handle shell built-in commands. The flaw enables sandbox bypass and remote code execution even with an empty command allowlist. The vulnerability allows attackers to manipulate environment variables through implicitly trusted shell built-ins like export, typeset, and declare. This...</summary> </entry> </feed>
