Claude Security adds continuous repository scans in public beta
Original title: Claude Security Now Scans Repositories Continuously to Find Vulnerabilities and Track Fixes
Enterprise customers get automatic rechecks that mark fixes and reopen recurring findings, with no fixed scan schedule.
Anthropic updated Claude Security so connected repositories are rechecked as they change rather than only when a person starts a scan. After a baseline deep scan, later checks run at intervals based on risk and complexity. The service can mark a finding fixed when a correction is detected and reopen it if the issue returns. It supports the Claude GitHub app, including GitHub Enterprise Server, plus tokens for GitLab and other Git providers, and adds an API, webhooks, and an MCP connector. It remains in public beta for Claude Enterprise customers.
Unconfirmed: The claim of more than one million vulnerabilities comes only from Anthropic and has not been independently confirmed.
- Anthropic says users found more than one million vulnerabilities since April; that figure is unconfirmed.
- Analysis uses Claude models, including Claude Mythos, without Cyber Verification Program access.
- A project can cover multiple repositories so cross-service trust issues can be assessed.
- Documentation says results can vary between runs and proposed fixes should be reviewed.
- Vendors
- Anthropic
- Products
- Claude SecurityClaude Code
- AI models
- Claude Mythos
From the original article
3 min read at cybersecuritynews.com
Anthropic has updated Claude Security so it can keep checking connected code repositories as they change, rather than waiting for a person to start each scan. The service also tracks reported vulnerabilities after developers make changes, marks them fixed when a correction is detected, and reopens them if they return. The updated product is in […]
By Guru Baran · cybersecuritynews.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.