タグ
説明
- name:quality-gates
- description:Python code quality gates script for linting, type checking, complexity analysis, and testing before commits. Use when enforcing consistent code quality across Python projects with fast (unit-tests) or comprehensive (all-tests) modes.
- license:Complete terms in LICENSE.txt
- metadata:[object Object]
Quality Gates
Overview
Project-agnostic bash script enforcing Python code quality gates using pyrefly, radon, hypothesis, pytest, and markdownlint. Three execution modes provide flexibility for different workflows: unit-tests for fast feedback, all-tests for comprehensive pre-merge validation, no-tests for static analysis only. Coverage runs are excluded by default; run coverage only when the user explicitly requests it.
When to Use
Before committing Python code to enforce consistent quality checks across projects without directory structure assumptions.
Modes
| Mode | Gates Run | Use Case |
|---|---|---|
| unit-tests | pyrefly, radon, hypothesis checks, pytest (unit/), markdownlint | Fast feedback during development |
| all-tests | pyrefly, radon, hypothesis checks, pytest (unit/ + integration/), markdownlint | Pre-merge/deploy comprehensive check |
| no-tests | pyrefly, radon, hypothesis checks, markdownlint | Static analysis only (time critical) |
Quick Start
Running quality gates:
Make sure uv is installed.
# Fast feedback: lint + type check + unit tests
.claude/skills/quality-gates/scripts/quality-gates.sh unit-tests
# Pre-merge: everything including integration tests
.claude/skills/quality-gates/scripts/quality-gates.sh all-tests
# Quick static checks only
.claude/skills/quality-gates/scripts/quality-gates.sh no-tests
Implementation Details
- Script:
scripts/quality-gates.sh- Uses uv/uvx for tool management - Excluded directories: tests, test, docs, doc, examples, scripts, build, dist, .venv, venv, .tox, .git, node_modules, pycache, *.egg-info
- Graceful fallbacks: Skips missing tools without failing
Common Anti-Patterns
| Rationalization | Reality |
|---|---|
| "Time-critical, skip checks" | Use no-tests mode for fast static validation |
| "Already tested manually" | Automation catches edge cases manual testing misses |
| "Partial checks sufficient" | Modes provide full coverage options; incomplete coverage misses bugs |
| "Coverage by default" | Coverage pollutes context; run it only when explicitly requested |
| "Too slow" | Tools are fast; skipping misses complexity and type issues |
| "Script not in project" | Copy to project to .claude/; run locally |
| "Tools not installed" | Install uv; graceful fallbacks skip missing tools |
Red Flags
Stop and re-evaluate if you're:
- Skipping all checks
- "This time is different"
- "Good enough for now"
- Rationalizing under deadline pressure
- Adding coverage without an explicit request
All of these suggest running the script with the appropriate mode.
関連 Skill
興味のある他の Skill を探索
requesting-code-review
Use when completing tasks, implementing major features, or before merging to verify work meets requirements
verification-before-completion
Use when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always
mcp-builder
Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).