CVE-2026-21852
Claude Code is an agentic coding tool. Prior to version 2.0.65, vulnerability in Claude Code's project-load flow allowed malicious repositories to exfiltrate data including Anthropic API keys before users confirmed trust. An attacker-controlled repository could include a settings file that sets ANTHROPIC_BASE_URL to an attacker-controlled endpoint and when the repository was opened, Claude Code would read the configuration and immediately issue API requests before showing the trust prompt, potentially leaking the user's API keys. Users on standard Claude Code auto-update have received this fix already. Users performing manual updates are advised to update to version 2.0.65, which contains a patch, or to the latest version.
5 articles across 4 outlets · first covered May 7, 2026 · latest May 7, 2026
Coverage timeline
-
One‑click TrustFall exploit hijacks Claude Code via bad repowww.darkreading.com · May 7, 2026
-
Anthropic’s Claude Code flaws let repos seize PCs, steal keyswww.darkreading.com · Feb 25, 2026
-
Anthropic's Claude Code AI exploited by repos to steal API keyssecurityaffairs.com · Feb 25, 2026
-
Anthropic's Claude Code AI bugs let attackers run code and steal API keysthehackernews.com · Feb 25, 2026
-
Check Point finds API key leak in Anthropic's AI code assistantresearch.checkpoint.com · Feb 25, 2026