The Agent Skills Directory

[SAFE]: The skill instructions promote security best practices by requiring adversarial testing, input validation, and evidence-based verification of code changes.
[COMMAND_EXECUTION]: The run-checks.sh script automates execution of standard development tools (like npm test, pytest, cargo test). It uses hardcoded command templates based on local project file detection, which is standard and safe behavior for a verification tool.
[DYNAMIC_EXECUTION]: The skill dynamically identifies the appropriate verification commands for a project at runtime using a rule-based detection script, which poses no security risk in this context.

verifying-implementations