Claude Security now scans as code changes and tracks fixes
Anthropic added continuous repository scanning, automatic fix tracking, team-informed threat models, and API/MCP integrations to Claude Security.
AT A GLANCE
Key points
- Continuously scans repositories as code changes
- Reopens findings if fixed issues return
- Uses project knowledge and editable threat models
- Connects to workflows through API, webhooks, and MCP
What is changing?
Anthropic released the updated Claude Security on October 21 as a public beta for Claude Enterprise customers. It performs a baseline scan of connected GitHub or GitLab repositories, then regularly reviews code changes according to risk and complexity. A project can include multiple repositories, allowing the system to look for security issues that span services.
Open findings are checked against changed code and marked fixed when the repair ships; if the issue appears again, the finding reopens. Teams can add deployment details and security policies to a project knowledge base. Claude’s editable threat model helps determine severity and explain possible attack paths.
A new API and webhooks connect the product to existing workflows, while an MCP connector lets engineers inspect findings, start fixes, and resolve results inside Claude.ai and Claude Code. Anthropic says the product has found more than one million vulnerabilities across thousands of organizations.
Explore the original sources
Source date: 21 October 2026.
Published by AI Radar: 22 Oct 2026, 03:06 (Istanbul, UTC+3).
FOLLOW THE TOPIC
Topics and tags
Share your discovery.
Send this story to a friend.
Reader comments
What do you think about this development? Share your experience or question. Comments are published after editorial approval.
Your name/nickname and comment become public if approved. Read: Privacy notice.
Loading comments…