coding-effectively_skill

Overview

This skill enforces disciplined, context-aware coding practices to improve correctness, maintainability, and cross-platform reliability. It always applies core sub-skills (functional vs imperative separation and defense-in-depth) and selects additional sub-skills based on language, runtime, and task. Use it for writing, refactoring, reviewing, or designing features across projects and stacks.

How this skill works

The skill inspects code and design decisions against core principles: property-driven design, correctness-first error handling, pragmatic incrementalism, and clear file/module organization. It flags risky patterns (catch-all utils, swallowed errors, premature abstractions, platform mixing) and recommends concrete fixes and tests. When applicable, it activates conditional sub-skills for TypeScript, React, Postgres, and testing strategies.

When to use it

• Writing new features or refactoring existing code
• Designing data transformations or APIs where invariants matter
• Performing code reviews focused on correctness and error handling
• Preparing cross-platform or multi-environment deployments
• Creating or improving test suites, including property-based tests

Best practices

• Model full error space: differentiate user-facing and internal errors with low-ceremony messages
• Separate pure logic from side effects and validate data at every boundary
• Prefer concrete, composable code and delay abstraction until pattern appears three times
• Name files by purpose (string-formatting.ts) instead of generic utils.ts
• Keep platform-specific code in dedicated modules and document differences
• Use the type system to encode correctness; avoid any/type-bypasses

Example use cases

• Designing a serializer where roundtrip and invariants must be proven and tested
• Refactoring a large helper file into purpose-named modules for discoverability
• Reviewing error handling to ensure actionable, composable user-facing messages
• Adding platform-specific implementations for process spawning with separate files
• Choosing property-based tests for pure functions and serialization logic