The Agent Skills Directory

[SAFE]: The skill implements a dedicated security module (generator/security.py) providing path validation (validate_project_root) to prevent directory traversal attacks during the project analysis phase.
[SAFE]: It includes protection against Denial of Service (DoS) attacks, such as JSON bombs, by enforcing file size limits during JSON parsing operations.
[SAFE]: File reading and writing operations are centralized and use path resolution to ensure all activity remains within the user's project directory.
[SAFE]: Test generation uses a safe, string-based template system (str.format()) rather than dynamic execution functions like eval() or exec(), preventing arbitrary code injection during the rendering phase.
[SAFE]: Credentials found in generated fixtures and documentation are clearly marked as placeholders for testing purposes and are tagged with ignore-patterns for security scanners.
[SAFE]: The skill performs static analysis of the project structure and source code to detect frameworks and routes, avoiding the execution of the application's code.

e2e-outside-in-test-generator