dry-wit

Use when maintaining Bash code and the next move should be a fast Bats regression loop that narrows failures before editing more code.

Use when a Bash maintenance task needs coordinated testing, linting, and benchmarking instead of treating those loops as separate afterthoughts.

Use when Bash maintenance work needs a deliberate plan for which untested behaviors to cover next, especially after failures or shared-module edits.

Use when Bash performance work should be measured with repeatable hyperfine benchmarks instead of intuition or one-off timing.

Use when Bash maintenance work should be driven by ShellCheck findings, especially when warnings need prioritization and safe fixes.