On January 14, 2026, Pillar Security disclosed CVE-2026-22708, a high-severity flaw in Cursor that allowed shell built-ins like export, typeset, and declare to execute without appearing in the allowli…
Pillar Security researchers identified an active campaign, dubbed Deadbugz, that distributes a malicious Model Context Protocol (MCP) server named 'productivity-suite' via 23 GitHub pull requests from…
Security researchers from Pillar Security discovered vulnerabilities in Google's Agent Development Kit for Python that could allow malicious actors to manipulate AI agents into triggering unauthorized…
Pillar Security researchers demonstrated an agent-to-agent privilege boundary failure in Google's Agent Development Kit (ADK) repository, where a low-privilege triage agent could be manipulated via pr…
Pillar Security discovered multiple privilege escalation flaws in Google's Agent Development Kit for Python, which has over 90 million downloads, allowing public-facing AI agents to trigger more privi…
Pillar Security reported that security flaws in Google's Agent Development Kit for Python could allow public-facing AI agents to trigger privileged automation, enabling manipulation of pull-request re…
Pillar Security researchers disclosed on August 3, 2026, a novel attack against Google's open-source Agent Development Kit for Python, specifically targeting workflows in the gemini-cli GitHub reposit…
Pillar Security disclosed on August 3 two now-patched privilege-boundary weaknesses in Google's Agent Development Kit for Python repository, one allowing prompt injection to cross from a public triage…
Pillar Security's research demonstrates that AI coding agents can be manipulated to bypass their sandbox through trusted files and tools, posing a security risk in CI/CD pipelines. The researchers sho…
Pillar Security found that four top AI coding agents — Cursor, OpenAI's Codex, Google's Gemini CLI, and Antigravity — can be made to cross their security boundaries via prompt injection, without techn…
Pillar Security disclosed vulnerabilities showing AI coding agents in Cursor, Codex, Gemini CLI, and Antigravity can indirectly cross security boundaries without escaping their sandboxes. The research…
Security researchers at Pillar Security broke out of the sandboxes in four widely used AI coding agents — Cursor, OpenAI's Codex, Google's Gemini CLI and Antigravity — by having the agent write a file…