A security researcher has demonstrated a critical vulnerability affecting AI coding agents like Anthropic's Claude Code and Google's Gemini CLI. By opening a specially crafted GitHub issue, an attacker can gain remote code execution on the vendor's own CI/CD infrastructure, potentially stealing API keys and other sensitive information. This repeatable attack pattern was presented at Black Hat USA and affects the default configurations of these AI coding assistants, with Anthropic and Google having released patches. AI
IMPACT Exposes a significant trust gap in AI coding assistants, potentially leading to widespread compromise of developer workflows and sensitive data.
RANK_REASON Security vulnerability disclosure affecting AI-powered developer tools.
Read on dev.to — Claude Code tag →
- Anthropic
- Claude Code
- CVE-2026-12537
- CVE-2026-54316
- Elad Meged
- Gemini CLI
- GitHub
- Novee Security
- OpenAI
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
