Claude Code's “Don't Ask Again” File Exposes Passwords and Keys in 1,122 Public Copies on GitHub
Over 1,100 public GitHub repos contain plaintext credentials persisted by Claude Code's approval-caching feature—a novel secret-sprawl vector that bypasses traditional pre-commit scanners and demands new AI-tool governance policies.
Summary written by editorial AI · Source link below
Passwords and API keys that developers typed into commands they approved in Claude Code are sitting in public GitHub repositories, where anyone can read them.
When a user selects "Yes, and don't ask again," Anthropic's AI coding assistant can save the command, word-for-word, including the secret, in a project settings file.
Software security company Apiiro said yesterday that 1,122 public
Editorial Analysis
AI coding assistants introduce a credential-leakage path that existing secrets-management toolchains were not designed to catch, expanding the attack surface for any organisation adopting these tools.
Mandate that AI coding assistant configuration and approval-cache files are added to .gitignore and scanned by secret-detection tools before any code reaches a remote repository.
AI coding tools are inadvertently publishing developer credentials to public repositories, creating a new category of credential-exposure risk.
Forward-looking interpretation drafted by editorial AI under human review — not a reproduction of the source. See methodology.
External link — opens at THN (Feedburner) in a new tab.
More from the AI Security Desk
- Runaway AI hacking agent behind WorkOS breach after escaping sandbox3d
- AI Gave Your Incident Archive a Second Job3d
- Malicious Browser Extensions Could Use Claude Dispatch to Run Code on Users' Computers3d
- AI Slop and Hallucinations in Vulnerability Assessment: A Survey on Reasoning Failures and Trustworthy Mitigation3d
- Persona-Model Collapse in Emergent Misalignment3d