plugins/tech-lead/skills/code-review/SKILL.md
# Code Review Skill This skill should be used when the user asks for "code review", "review my changes", "review this PR", "check my code", "pre-merge review", "review diff", or mentions reviewing code quality, implementation correctness, or preparing changes for merge. ## Overview Code review following the **Review Pyramid** methodology — a prioritized approach that focuses on what matters most. ## The Review Pyramid ``` ▲ /|\ 5. Code Style (Nit) ← Least important,
npx skillsauth add ruslan-korneev/claude-plugins plugins/tech-lead/skills/code-reviewInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
This skill should be used when the user asks for "code review", "review my changes", "review this PR", "check my code", "pre-merge review", "review diff", or mentions reviewing code quality, implementation correctness, or preparing changes for merge.
Code review following the Review Pyramid methodology — a prioritized approach that focuses on what matters most.
▲
/|\ 5. Code Style (Nit) ← Least important, automatable
/ | \
/ | \ 4. Tests
/ | \
/ | \ 3. Documentation
/ | \
/ | \ 2. Implementation Semantics ← High priority
/ | \
/________|________\ 1. API Semantics ← Most critical, review first
Key insight: Review from bottom to top. API and implementation issues are far more important than style issues.
Fast first pass using quick-reviewer agent (sonnet):
Thorough analysis using code-reviewer agent (opus):
| Trigger | Action |
|---------|--------|
| "Review my staged changes" | /review:diff staged |
| "Review before I merge" | /review:diff main --mode=full |
| "Quick check my code" | /review:diff --mode=quick |
| "Deep review this feature" | /review:diff main --task="..." --mode=deep |
--task flag for accurate business logic reviewpyramid.md — Detailed explanation of the Review Pyramidapi-semantics.md — API review checklistimplementation.md — Implementation review checklistsecurity.md — Security review checklisttesting.md — Test review checklisttools
# Architecture Patterns System-level architecture patterns for Python backend projects (FastAPI + SQLAlchemy 2.0). ## Triggers Use this skill when the user: - Designs a new system or major feature - Asks about project structure or module boundaries - Makes data modeling decisions - Designs API contracts - Evaluates architectural trade-offs - Creates or reviews Architecture Decision Records (ADR) - Needs to modernize legacy code ## Abstraction Levels | Plugin | Level | Focus | |--------|----
tools
# Python Typing Patterns Python type annotation patterns without `type: ignore`. Always the correct solution. ## Triggers Use this skill when the user: - Gets mypy/pyright errors - Asks about Python type annotations - Wants to add type hints - Works with generics, protocols, TypeVar ## Main Principle: NEVER type: ignore Every type error has a correct solution. `type: ignore` is: - Masking potential bugs - Disabling type checking - Technical debt More details: `${CLAUDE_PLUGIN_ROOT}/skills/
tools
# Ruff Patterns Knowledge about the ruff linter and patterns for solving errors. **ZERO noqa policy** — always look for the proper solution. ## Triggers Use this skill when the user: - Asks "how to configure ruff" - Gets a ruff error and wants to understand how to fix it - Wants to migrate from black/isort/flake8/pylint - Asks about a specific rule (E501, F401, etc.) ## Main Principle: NEVER USE NOQA **Every ruff error has a proper solution.** Using `# noqa` is: - Hiding the problem, not so
tools
# Pytest Patterns Quality testing patterns with pytest for FastAPI projects. **TDD-first approach**. ## Triggers Use this skill when the user: - Wants to write tests - Asks about pytest, fixtures, mocks - Wants to set up testing for FastAPI - Uses TDD approach ## Main Principle: TDD 1. **Red** — write a failing test 2. **Green** — write minimal code 3. **Refactor** — improve while keeping tests green ## What Makes a Good Test ### 1. Clear Name ```python # ✅ Good — describes what, when, a