Clean code

Clean Code Analyzer

Run the clean-code static analyzer on the targeted Python file or directory to enforce formatting, structure, and readability standards.

Workflow

1. If target_path is provided: Execute: clean-code check "{{target_path}}"

2. If no target_path is provided but an instruction is:

   • Generate the requested Python code based on the instruction.
   • Save the code to a temporary path ({{generated_file_path}}).
   • Execute: clean-code check "{{generated_file_path}}"

Guidelines

• By default, info-severity violations (the fuzziest, lowest-signal rules) are hidden to avoid context noise.
• If the summary indicates violations hidden below --min-severity warning, and a deeper pass is required, rerun the command with: clean-code check --min-severity info "{{target_path}}" (or the generated path).
• Review the output from the analyzer. If any violations are reported, modify the source code directly using the suggested fix: patterns.
• Do not silence violations using inline comments unless specifically requested by the user.
• DP701/SD801/SD802 findings are SOLID/DRY smells, not local style nits — the fix: suggestion (extract a shared helper, replace a type-switch with polymorphism/a dispatch table, split a class) usually means introducing a new function/class or changing call sites, not a one-line edit. Apply the refactor, but summarize what you restructured and why when presenting the result, since these findings reshape the code's structure more than the other rules do.
• DP701 compares functions across every file reached in one check run — it still runs (and can still catch within-file duplicates) on a single-file target_path, but a single-file check can't see a duplicate living in a sibling file. Prefer checking the containing directory (or the whole project) when hunting for cross-file duplication, not just the one file you touched.
• Present the final cleaned code back to the user.