Vulnerability intelligence
CVE-2026-22708
Tenet Security disclosed a new attack class on June 12. Attackers inject prompts into Sentry error events using publicly discoverable DSNs. AI coding agents retrieve the events via MCP and execute attacker-controlled code. Sentry called it 'technically not defensible.'
2026
What WebPulse reported · 2 analyses
Agentjacking: A Malicious Sentry Error Event Can Hijack Claude Code, Cursor, and Codex. 2,388 Organizations Exposed.
Tenet Security disclosed a new attack class on June 12. Attackers inject prompts into Sentry error events using publicly discoverable DSNs. AI coding agents ret
June 15, 2026
Claude Code GitHub Action Had a Prompt Injection Flaw. A Malicious Issue Title Could Read Your CI/CD Secrets.
CVE-2026-22708, CVSS 7.8. A crafted GitHub issue description caused Claude Code's GitHub Action to read CI/CD secrets from /proc/self/environ. Patched in v1.0.9
June 14, 2026
Related vulnerabilities