project-development-skill

Shared vocabulary for designing deep modules — depth, seams, adapters, testability. Inspirado por mattpocock/skills.

Resolver conflitos de merge/rebase de forma estruturada. Inspirado por mattpocock/skills.

Disposable experiments to validate an idea before committing to a build. Validate feasibility, compare approaches, surface unknowns. Also covers quick prototypes to answer a single design question (logic or UI).

Evolve prompts, code, and configurations through reflective optimization with AI feedback.

Regression testing patterns for AI-assisted development. Sandbox/production path parity, response shape contracts, data completeness, and patterns to catch AI blind spots.

REST API design patterns — resource naming, HTTP semantics, status codes, versioning, pagination, error handling, rate limiting, and OpenAPI specs. Use when designing or reviewing REST APIs.

Guard your API: JWT, OAuth 2.0, RBAC, sessions, and API key patterns for secure authentication and authorization.

Use when writing, reviewing, or refactoring code for readability and maintainability. Covers naming, functions, comments, error handling, classes, testing, and common code smells with concrete rules and anti-patterns.

Database patterns — migrations, indexing, N+1 prevention, connection pooling, transactions, repository pattern, CQRS, soft delete, audit trails, and query optimization. Use when designing or working with database layers.

Use when designing or refactoring code with Domain-Driven Design patterns. Covers strategic (Bounded Contexts, Context Maps) and tactical (Entities, Value Objects, Aggregates, Repositories, Services, Events, Modules, Factories) design with decision trees and code examples.

Deployment patterns — blue-green, canary, rolling updates, feature flags, CI/CD, container orchestration. Use when designing deployment pipelines, release strategies, or production operations.

Design patterns reference — GoF 23 patterns, decision trees, anti-patterns, and practical examples. Use when designing architecture, choosing abstractions, or refactoring code.

Documentation patterns — README structure, API docs, ADRs, code comments, changelogs, contributing guidelines. Use when writing, organizing, or improving project documentation.

Monitoring and observability — structured logging, metrics, distributed tracing, health checks, alerting, dashboards, and SLI/SLO/SLA concepts. Use when setting up observability for applications or services.

Bomb-proof development pipeline: brainstorm → spec → plan → code → test → review. Master orchestrator for software projects.

Performance optimization patterns — caching, lazy loading, profiling, memory management, database queries, frontend performance. Use when optimizing speed, reducing resource usage, or fixing performance bottlenecks.

Bomb-proof plan mode: write an actionable markdown plan, no execution. Bite-sized tasks, exact paths, complete code.

Practical recipes combining multiple skills — create APIs, debug production issues, refactor legacy code, add authentication, set up projects, and optimize performance. Each recipe is a step-by-step guide with exact commands and expected outputs.

Fresh-eyes verification: pre-commit security scan, quality gates, two-axis review, auto-fix loop.

Security checklist covering OWASP Top 10, input validation, authentication bypass, secrets management, CORS, CSP, injection prevention, XSS, CSRF, rate limiting, password hashing, session management, and dependency scanning. Use when building or auditing secure applications.

Wire services together: Worker-Function-Trigger primitives for event-driven, agentic, and distributed systems.

4-phase root cause debugging: understand bugs before fixing.

Red-first: write the failing test, watch it fail, then code. Enforces RED-GREEN-REFACTOR with no shortcuts.

Remove signs of AI-generated writing from text. Detects and fixes 24 patterns including inflated symbolism, promotional language, superficial -ing analyses, em dash overuse, rule of three, AI vocabulary words, and excessive hedging. Based on Wikipedia's comprehensive "Signs of AI writing" guide.

Fresh-start execution: dispatch a clean subagent per task with two-stage review (spec then quality).

Write clear, forceful prose for humans. Based on Strunk's Elements of Style. Use for documentation, commit messages, error messages, reports, UI text, or any writing a human will read. Covers active voice, omitting needless words, concrete language, and AI writing patterns to avoid.

Write bomb-proof implementation plans: bite-sized tasks, exact paths, complete code, verification steps. Use before every multi-step implementation or delegation.