Comment and Control: Claude Code Security Review leaking ANTHROPIC_API_KEY via prompt injection

Comment and Control: Prompt Injection to Credential Theft in Claude Code, Gemini CLI, and GitHub Copilot Agent

How GitHub comments exposed credentials in Anthropic Claude, Google Gemini, and Microsoft Copilot workflows. Includes reproductions, vendor responses, and the ACM CCS 2026 paper.

April 15, 2026 · 16 min · 3233 words · Aonan Guan