Claude Security rescans as code changes and closes findings when fixes ship, replacing scheduled scans
Synopsis
Claude Security has moved from on-demand scanning to continuous coverage: after a repository is connected it runs a deep baseline scan, then rescans changes according to their risk and complexity, and marks findings fixed when the fix lands. Previously a scan captured one moment of the code, leaving teams to schedule scans themselves. Each project also carries a knowledge base and an editable threat model that shape what gets flagged.
Interpretation
Claude Security now keeps up with repositories as code changes and tracks which findings get fixed, so teams no longer schedule scans themselves. A scan captures one moment while code changes every day, and teams asked for coverage that stays current without anyone scheduling it. Since entering public beta in April, thousands of organizations have used it to find over a million vulnerabilities; the latest version is available today in public beta for Claude Enterprise customers.
Setup means creating a project and connecting repositories, after which a deep baseline scan runs and changes are scanned regularly based on their risk and complexity. Repositories connect through the Claude GitHub app, including GitHub Enterprise Server, or with an access token from GitLab and other providers, with a branch and scope chosen for each. A project can hold multiple repositories, so issues between services can be caught, such as a backend that trusts input a client was supposed to validate.
Whether a finding counts as fixed is decided by the code change rather than by the tool that found it. Most fixes happen outside the tool that found the issue, whether the team fixed it in Claude Code or in a separate process. Open findings are checked against the code that changed and marked fixed when the fix lands; if the issue is found again later, the finding reopens automatically.
Perspective
The report covers the feature set described for the latest Claude Security version: continuous scanning, fix tracking, a per-project knowledge base and editable threat model, severity and attack-path explanations, an API and webhooks, an MCP connector, and role-based access controls. The version is available in public beta for Claude Enterprise customers, and admins enable it in the admin console and set who can access it. The intended readers are engineering and security teams responsible for code scanning and triage.
The announcement gives no detection rate, false-positive rate, or comparison with other scanning tools, and the usage scale and customer statements are the vendor's own account. The cadence of risk- and complexity-based rescans is not specified. The free allowance is $25,000 per month for two months, applying to usage through December 31, 2026.