Enforce readable, maintainable code with naming conventions, function size limits, and cognitive complexity guidance.
Structured checklist covering correctness, security, performance, and readability.
Consistent error handling strategies with proper logging and user-facing messages.
Project-wide naming rules for files, functions, variables, constants, and types.
Safe, incremental refactoring patterns with clear criteria for when to refactor vs rewrite.
Applied SOLID with concrete examples tailored to your language and architecture.
Don't Repeat Yourself and You Aren't Gonna Need It — preventing over-engineering.
JSDoc/docstring standards, inline comment rules, and README structure.
Enforce conventional commit format enabling automated changelogs and semantic versioning.
Consistent branch naming patterns that integrate cleanly with issue trackers and CI/CD.
PR size limits, description templates, review etiquette, and merge strategies.
SemVer rules and communicating breaking changes clearly.
Trunk-based development with feature flags and short-lived branches.
Arrange-Act-Assert, test naming, isolation principles, and coverage targets.
Testing service interactions, database operations, and API contracts.
End-to-end testing with Playwright: selectors, wait strategies, CI configuration.
Red-Green-Refactor cycle and how to integrate AI into the TDD workflow.
When to mock vs. use real implementations, and avoiding over-mocking.
Meaningful coverage targets by code type and what coverage doesn't tell you.
When snapshots help vs. hurt for UI components.
Composition patterns, compound components, and when to use each approach.
Decision tree for local state vs. Zustand vs. TanStack Query.
App Router patterns: server/client components, layouts, route handlers, metadata API.
Tailwind config conventions, custom tokens, and component class patterns.
WCAG 2.1 AA compliance, semantic HTML, ARIA usage, and keyboard navigation.
Core Web Vitals targets, image optimisation, bundle analysis, lazy loading.
CSS Modules vs. Tailwind vs. CSS-in-JS decision guide and global style management.
shadcn/ui integration, component variants with cva(), and Storybook docs.
Strict TypeScript config, rules to enforce, and common strict mode pitfalls.
Discriminated unions, branded types, utility types that prevent bugs at compile time.
When to use generics, constraints, inference, and avoiding over-generic code.
Zod schema patterns for API inputs, form validation, and env variable parsing.
PEP 8+ rules via Ruff, Google-style docstrings, consistent project structure.
Full type annotation coverage with mypy strict mode.
asyncio patterns, FastAPI async conventions, and how to test async code.
Production-grade CLIs with Typer: commands, options, Rich output.
Modern Python dependency management with uv: lockfiles, virtual envs.
Resource naming, HTTP methods, status codes, pagination, and versioning.
Schema design, DataLoader for N+1, subscriptions.
JWT patterns, session management, RBAC, and authentication pitfalls.
Rate limiting strategies, response headers, user-friendly errors.
Prisma/SQLAlchemy conventions, N+1 prevention, migrations, transactions.
Consistent error response structure and machine-readable codes.
Receiving webhooks securely: signature verification, idempotency.
Multi-stage builds, image size optimisation, security hardening.
Workflows for lint, test, build, deploy with caching and matrix strategies.
.env conventions, environment configs, and secret rotation.
Structured JSON logging with pino/structlog, log levels, OpenTelemetry.
Vault, AWS Secrets Manager, GitHub OIDC for zero-trust secret handling.
Opinionated folder structure for different project types.
Turborepo configuration, workspace structure, shared packages.
In-memory, Redis, CDN, and HTTP caching with clean invalidation.
OWASP Top 10 coverage, dependency auditing, security review.
System prompts, few-shot examples, chain-of-thought, and output formatting.
Anthropic SDK patterns, streaming, error handling, retry logic, cost management.
Retrieval-Augmented Generation: chunking, embedding, retrieval, re-ranking.
ReAct pattern, tool design, agent memory architecture, multi-agent orchestration.
Building and integrating MCP servers with Claude Code.
Andrej Karpathy's core coding philosophy adapted for Claude Code workflows.