Claude Security now checks your code for flaws all the time

Claude Security, a tool from Anthropic that looks for security flaws in software code, no longer waits for someone to start a scan. It now keeps watch on a company's code as it changes, tracks each problem it finds until it is fixed, and reopens the problem if it comes back. The update is in public beta for Claude Enterprise customers.

What changed

A scan only captures one moment, and code changes every day. Teams asked for coverage that stays current without anyone scheduling it. Claude Security now runs a deep baseline scan first, then rescans changes regularly based on their risk and complexity. One project can hold several repositories, which are the folders where code is stored, so it can catch issues that sit between services.

It also checks open findings against new code and marks them fixed when the fix lands. That works whether the team fixed the issue in Claude Code or in another process.

Context and priorities

Each project now has a knowledge base where a team can describe how its code is deployed and which security policies apply. Claude Security also builds a threat model from the code, which is a map of what could go wrong. Teams can review and edit it. Later scans use it to decide what is worth flagging and to skip risks the environment already handles.

Each finding gets a severity rating judged against that threat model, plus an explanation of how an attacker could reach the weak code. Some include a diagram of the attack path.

Connecting to other tools

• A new API lets teams list findings, resolve or reopen them, and manage projects.
• Webhooks can alert other systems when a new finding appears.
• An MCP connector works in Claude.ai and Claude Code, so an engineer can review a finding, start a fix and mark it resolved in one session.
• New access controls let admins decide who can view projects, triage findings and create projects.

Cost and availability

Admins can turn it on in the admin console. Organizations get $25,000 in free usage credits each month for two months, for usage through December 31, 2026. Full terms are in the Help Center. Since the April beta launch, thousands of organizations have used the tool to find over a million vulnerabilities. Pricing after the free credits is not stated in the material.