🛡️ CVE-2026-21852

The items that name it 8

• Is Claude Code safe for the enterprise? What the official settings cover, and what they don't
  article2026-10-22It is written here as CVE-2026-21852: Claude Code project-level configuration loaded before the trust dialog resolved, which let an attacker exfiltrate API keys by pointing the agent at an attacker's endpoint. Fixed..
• Top 10 attacks on Claude Code and what each one actually broke
  writeup2026-09-23

  Anthropic API key theft via a settings file pointing ANTHROPIC_BASE_URL at an attacker endpoint; found by Check Point Research; fixed in Claude Code 2.0.65; patched before publication.

• OWASP Agentic Skills Top 10 explained: the ten agent skill risks, and which to fix first
  article2026-08-25

  Claude Code vulnerability where repository configuration files could trigger API token exfiltration before any dialog appeared; patched

• Vulnerability Garden
  tool2026-05-13
• anmolksachan/AI-ML-Free-Resources-for-Security-and-Prompt-Injection
  repo2026-05-07

  Claude Code API key exfiltration, CVSS 5.3, patched Jan 2026

• CVE-2026-21852 — Claude Code Information Disclosure (pre-trust config injection leaks API keys)
  advisory2026-01-21

  Claude Code Information Disclosure (pre-trust config injection leaks API keys); configuration injection flaw (CWE-522: Insufficiently Protected Credentials) in Claude Code's project-load flow allowing malicious repositories to exfiltrate Anthropic API keys before users confirm trust

• OWASP Agentic Skills Top 10 (AST10)
  repo

  Second vulnerability disclosed alongside CVE-2025-59536, both patched months earlier before public disclosure.

• OWASP Agentic Skills Top 10
  paper

  Disclosed alongside CVE-2025-59536 by Check Point Research; both were patched months earlier but disclosure confirms repository-controlled config files can silently execute shell commands and exfiltrate API keys.