The Agent Skills Directory

[SAFE]: The skill instructions and associated reference files were audited across all 10 threat categories, and no malicious patterns, safety bypasses, or security vulnerabilities were identified. The content is educational and serves as a technical style guide.
[PROMPT_INJECTION]: No prompt injection or behavior override attempts were detected. The instructions are directed at maintaining code quality and following established Python standards.
[DATA_EXFILTRATION]: No data exfiltration patterns or unauthorized network operations were found. The skill correctly identifies the risks of hardcoding secrets and promotes industry-standard practices for secret management.
[REMOTE_CODE_EXECUTION]: No suspicious remote code execution patterns were found. References to package managers and linting tools like uv and ruff are standard for the described development environment.
[COMMAND_EXECUTION]: No unauthorized or dangerous command execution instructions are present. Shell commands are mentioned only as illustrative examples for formatting and testing.
[CREDENTIALS_UNSAFE]: No hardcoded credentials or private keys were found. The skill emphasizes the secure use of environment variables for sensitive configuration.

best-practice-python