TL;DR
Get privacy and security gear delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
Apiiro says it found passwords in 1,122 public GitHub files associated with Claude Code’s “Don’t Ask Again” permission feature. The supplied report does not establish how many credentials were valid, whether they were accessed, or whether Anthropic has confirmed the findings.
Security company Apiiro says it identified passwords in 1,122 public GitHub files associated with Claude Code’s “Don’t Ask Again” permission setting, raising concerns about sensitive information being committed to public repositories during AI-assisted development. The material provided for this report does not establish whether the exposed credentials were still valid, accessed by others, or confirmed by Anthropic.
Apiiro’s report describes a pattern in which files tied to Claude Code’s “Don’t Ask Again” behavior appeared in public GitHub repositories alongside passwords. The headline figure is 1,122 files; it refers to files, not necessarily 1,122 distinct users, accounts, repositories, or unique passwords. The source material supplied here does not include the report’s underlying dataset, search method, or examples sufficient to independently verify the count.
The central security issue is that credentials can become publicly readable when included in source files or related project artifacts. A password or token committed to a public repository may be copied or indexed, even if the repository owner later removes it. But the available material does not say what types of accounts the passwords could access, whether the credentials remained active, or whether anyone used them.
Apiiro is a security vendor, and its finding should be understood as a company-reported result, not as an independently verified incident count. The supplied source contains no response from Anthropic, no confirmation from GitHub, and no details about remedial steps taken by repository owners.
Risks of Credentials in Public Repositories
If even some of the reported passwords were active, their appearance in public files could put associated accounts and systems at risk. Public repositories can be inspected by anyone, while copied credentials may persist beyond the removal of the original file. That makes prompt rotation and access review common defensive steps for anyone who discovers a live secret in a repository.
The finding also points to a broader concern for teams using coding assistants: permission and workflow features may affect what gets written into project files, but they do not replace secret-handling controls. Developers and organizations still need checks that prevent credentials from entering version control, and procedures for responding if a secret is exposed.
The scale and practical impact remain uncertain. A file count alone cannot show how many distinct secrets were exposed or whether any resulted in unauthorized access. Readers should not infer a confirmed breach from the reported number.
Products Worth Considering
Apple iPad 11-inch with A16 Chip, 512GB, Wi-Fi 6, and Apple Pencil (USB-C) — Blue
Bose QuietComfort Ultra Headphones (2nd Gen), Wireless Noise-Cancelling Over-Ear Headphones, Black
Logitech MX Keys S Combo Wireless Keyboard and Mouse, Graphite
As an affiliate, we earn on qualifying purchases.
Claude Code Permission Setting
Claude Code is Anthropic’s coding assistant. The “Don’t Ask Again” wording refers to a permission choice in the product, but the source material provided does not explain precisely which action the setting governs or how the relevant files were generated. It would be speculative to attribute the appearance of passwords to a particular product behavior without those details.
Apiiro’s report places the finding within the risk of secrets appearing in code hosted publicly on GitHub. A repository file may contain credentials intentionally or accidentally; the report’s headline links the files to Claude Code’s setting, but the supplied text does not provide enough technical detail to establish the pathway for each case. The distinction matters: association reported by a vendor is not, on its own, proof that a product defect caused every exposure.
““Claude Code’s ‘Don’t Ask Again’ exposes passwords in 1,122 public GitHub files.””
— Apiiro
Credential Validity and Exposure Scope
The supplied report material does not specify how Apiiro searched GitHub, how it determined that files were connected to the permission setting, or whether the 1,122 files were deduplicated across repositories. It also does not say whether the count represents distinct credentials or whether any files contained repeated or already-revoked passwords.
There is no information here confirming that the credentials were valid when found, that attackers accessed them, or that affected account holders were notified. The source material also includes no comment from Anthropic or GitHub and no account of fixes, takedowns, or an investigation. Those gaps prevent a firm conclusion about the cause, severity, and current status of the reported exposures.
Verification and Credential Response
The next useful developments would be technical details from Apiiro about its search criteria, validation process, and handling of duplicate files, along with any response from Anthropic or GitHub. Confirmation of whether affected credentials were active and whether they have been revoked would help clarify the potential impact.
For developers who find a password or token in a public repository, removing the file alone may not be enough: the credential should be treated as exposed and reviewed or rotated through the relevant service. The supplied source does not say whether Apiiro, Anthropic, or GitHub has issued specific guidance or contacted affected users, so those actions remain unconfirmed.
Key Questions
What did Apiiro report?
Apiiro reported finding passwords in 1,122 public GitHub files that it associates with Claude Code’s “Don’t Ask Again” permission setting.
Does the number mean 1,122 people were affected?
No. The reported figure counts files. The source material does not establish how many distinct users, repositories, accounts, or credentials those files represent.
Were the passwords confirmed to be active or used?
The material provided does not say whether the passwords were valid when discovered or whether anyone used them to access an account or system.
Has Anthropic confirmed the finding?
No response from Anthropic is included in the supplied source material. The reported count is attributed to Apiiro and is not independently verified here.
What should a developer do if a credential appears in a public repository?
Treat it as exposed: remove it from the repository and contact the relevant service or account administrator to revoke or rotate the credential. Deleting the file alone may not address copies already made.
Source: apiiro.com
