fleet-scanning-security

Automated GitHub Actions security scanning runs zizmor to detect misconfigurations and vulnerabilities in your workflows, producing a graded report. It validates workflow files after configuration changes or before releases to maintain security hygiene.

Scanning Claude agent configs for security issues uses AgentShield to analyze CLAUDE configuration files and agent definitions. It identifies potential risks and generates an A-F graded security report to help validate configurations before deployment.

Dependency scanning runs alongside GitHub Actions and Claude config checks using the optional Socket CLI. It integrates dependency analysis into the multi-tool security scan to produce a comprehensive graded report covering repository configurations and dependencies.

Automated security scanning generates an A-F graded report based on vulnerability counts from zizmor and AgentShield scans. It stores a local, untracked report file and provides a handoff summary for compliance verification and remediation tracking.

Run multi-tool security scans after updating CLAUDE configs, agent definitions, or GitHub workflows, before releases, and for periodic security hygiene checks. Scanning after configuration changes ensures vulnerabilities are caught early across complex repositories.

The automated scanning report uses an A-F grading scale based on vulnerability counts from AgentShield and zizmor scans. This counts-based grade provides a clear security posture indicator for repository configurations, workflows, and dependencies.