The Agent Skills Directory

Management of External Context: The skill processes user-provided context as part of its scanning logic. The instructions explicitly categorize this input as untrusted analysis data and include constraints to ensure it cannot override tool instructions or the core scanning workflow, which is a standard security design pattern.
Orchestration of Security Tools: The skill coordinates several integrated modules for discovery and validation. These components are used within their intended scope to generate security insights and follow the platform's security guidelines.
Artifact Integrity and Verification: The workflow requires that all canonical security reports and findings be verified on disk before a scan is marked as complete. This process helps ensure that the final security outputs are complete and consistent.

deep-security-scan