s-hiraoku

code-quality

0
0
# Install this skill:
npx skills add s-hiraoku/synapse-a2a --skill "code-quality"

Install specific skill from multi-skill repository

# Description

Run code quality checks (ruff, mypy, pytest) and optionally simplify code. This skill should be used when the user wants to check code quality, run linters, run tests, or simplify recently modified code. Triggered by /lint, /check, or /code-quality commands.

# SKILL.md


name: code-quality
description: Run code quality checks (ruff, mypy, pytest) and optionally simplify code. This skill should be used when the user wants to check code quality, run linters, run tests, or simplify recently modified code. Triggered by /lint, /check, or /code-quality commands.


Code Quality Check

This skill runs code quality tools (ruff, mypy, pytest) and optionally simplifies code.

Usage

/code-quality [options]

Options

  • --all or -a: Run on all synapse/ files (default: recently modified files only)
  • --simplify or -s: Also run code-simplifier agent after checks pass
  • --fix or -f: Auto-fix ruff issues with --fix flag
  • --test or -t: Run pytest after linting passes
  • --full: Run all checks including tests (equivalent to --fix --test)

Workflow

Step 1: Identify Target Files

If --all flag is provided:
- Target all files in synapse/ directory

Otherwise:
- Run git diff --name-only HEAD~1 to get recently modified files
- Filter to only .py files in synapse/ or tests/

Step 2: Run Ruff Linter

ruff check [files]

If --fix flag is provided:

ruff check --fix [files]

Report any errors found. If errors remain after fix, stop and report.

Step 3: Run Mypy Type Checker

uv run mypy [files]

Report any type errors found. If errors exist, stop and report.

Step 4: Run Tests (if --test or --full)

If --test or --full flag is provided AND ruff/mypy passed:

pytest

For running specific tests related to modified files:

pytest tests/ -v

Report any test failures.

Step 5: Run Code Simplifier (Optional)

If --simplify flag is provided AND all previous checks passed:

Use the code-simplifier:code-simplifier agent to simplify the recently modified code:

Task tool with subagent_type: code-simplifier:code-simplifier
Prompt: Simplify and refine the recently modified code in [files].
Look for opportunities to reduce duplication, simplify conditionals,
and improve readability while maintaining all functionality.

Step 6: Report Results

Summarize:
- Number of files checked
- Ruff status (pass/fail, errors fixed if --fix)
- Mypy status (pass/fail)
- Pytest status (pass/fail, skipped if not run)
- Code simplifier status (if run)

Error Handling

When errors are found:

  1. Ruff errors: If --fix is provided, attempt auto-fix first. Report remaining errors with file:line format.
  2. Mypy errors: Report type errors with file:line format. Suggest fixes based on error messages.
  3. Test failures: Report failed test names and assertion errors. Do NOT proceed to simplification if tests fail.

Examples

Basic check on recent changes

/code-quality

Check all files with auto-fix

/code-quality --all --fix

Full quality check with tests

/code-quality --full

Run with tests only

/code-quality --test

Full quality check with simplification

/code-quality --simplify --test

Shorthand

/lint          # Same as /code-quality
/check         # Same as /code-quality
/lint -s       # With simplification
/lint -a -f    # All files with auto-fix
/lint -t       # With tests
/lint --full   # Full check (--fix --test)

Tool Configuration

This project uses the following configurations (from pyproject.toml):

Ruff

  • Target: Python 3.10
  • Line length: 88
  • Rules: E, F, I, UP, B, SIM
  • Ignored: E501, B008
  • Excluded: synapse/proto/a2a_pb2*.py

Mypy

  • Strict mode with disallow_untyped_defs
  • Tests have relaxed rules (ignore_errors)
  • Proto files excluded

Pytest

  • asyncio_mode: auto
  • Run with: pytest or pytest tests/ -v

# Supported AI Coding Agents

This skill is compatible with the SKILL.md standard and works with all major AI coding agents:

Learn more about the SKILL.md standard and how to use these skills with your preferred AI coding agent.