MR-Axel
User16 open skills for Claude Code: development (project-setup, dev, test, review, deploy, ship), marketing, and remote job search for LatAm. Config lives in your repo, nothing hardcoded inside.
Categories
Indexed Skills (15)
deploy
Run this repository's release pipeline: build, migrate, deploy, commit and push, using the commands and permissions in .claude/project-profile.md. Stops on the first failure and never pushes broken code. Use for: deploy, ship it, release, /deploy.
product
Review a change from the product and business side: does it serve the user need, are the tier gates right, are the states complete, is it discoverable. Reads your product model from .claude/project-profile.md. Use before shipping. Use for: product review, does this make sense, /product.
review
Code review for security, performance and quality on a diff, with every finding anchored to a real file and line. Use before shipping. Use for: review this, code review, check my changes, /review.
ship
Full delivery pipeline for one feature: plan, implement, validate, self-review, product check, then deploy with your approval. Chains the dev, test, review, product and deploy skills. Use for: ship this feature, end to end, full pipeline, /ship.
ux
Audit UI for responsiveness, accessibility, design system consistency and interaction states, against the design system in .claude/project-profile.md. Use for: UX review, audit this component, accessibility check, /ux.
community-manager
Senior community manager and growth marketer. Generates ready-to-publish social posts, threads, carousels and editorial calendars for X, LinkedIn, Instagram and others, built on your own brand profile and funnel. Use for: write a post, social content, content calendar, launch campaign, community manager, /cm.
decision-log
Keeps a living log of the non obvious development decisions, the ones you cannot recover by reading the code or the git history: why this model, why this limit, why this approach and not the other one. Writes the entry in the same commit that implements the decision, and reads the log before proposing something that contradicts one. Use for: por que hicimos esto asi, decision log, ADR, architecture decision, documentar una decision, DECISIONS.md, /decision-log, esto ya lo habiamos decidido.
deep-review
Interactive architecture review. Works through architecture, code quality, tests and performance, presenting each issue as options with effort, risk and impact so you decide the priority. Use for: architecture review, deep review, tech debt review, /deep-review.
deploy-qa
Guided QA after a deploy. Reads the diff that shipped, maps changed paths to the areas they can break, decides the depth (smoke, selective regression, full regression) from blast radius instead of habit, and walks a concrete test list. Starts by proving the new build is actually live, because a deploy that reports success can still be serving the old bundle. Use for: que pruebo despues del deploy, deploy QA, smoke test, verificar deploy, post deploy check, /deploy-qa, se rompio algo con el push, que testeo de este cambio.
design-system
Establish a design system where there is none, and enforce the one that exists so violations fail a check instead of surviving code review. Covers token layers, the traps that make tokens quietly wrong (alpha over variables, contrast by threshold, a token that cannot express a state), and the four gates that keep drift out. Use for: design system, design tokens, tokens, theme, dark mode, light mode, colores del sistema, hardcoded colors, /design-system, esto no respeta el design system, unificar estilos.
dev
Implement a feature or fix a bug following this repository's own conventions, read from .claude/project-profile.md rather than assumed. Use for: implement, build a feature, fix this bug, add support for, /dev.
feature-agents
Runs a feature through five specialist subagents instead of one context: spec, planner, tester, QA, deployer. Each gets only the tools its job needs, so the spec writer physically cannot write code and the planner physically cannot run commands. For changes big enough that phase discipline stops holding on its own. Use for: delegar por fases, subagentes, agentes especializados, spec agent, planner agent, /feature-agents, no empieces a codear todavia, quiero el plan antes de tocar codigo.
latam-job-search
Remote job search for LatAm-based candidates. Runs an onboarding interview to capture location, work modality, timezone tolerance, compensation floor, language levels and hard exclusions, then searches LinkedIn and freehire, verifies each posting against those gates before scoring it, and returns a deduplicated daily digest. Use for: buscar trabajo, búsqueda laboral, find jobs, job search, remote jobs, LatAm jobs, daily job digest, evaluate a job posting, is this job a fit, /job-search, /job-setup.
project-setup
Interviews you about a codebase and writes .claude/project-profile.md, the shared config that the dev, test, review, deep-review, ux, deploy, ship and product skills all read. Run once per repository. Use for: project setup, configure skills, set up my project, /project-setup.
linkedin-personal-brand
Sistema de marca personal en LinkedIn. Define identidad y voz, arma calendario de publicación, rastrea noticias de tu nicho y devuelve textos listos para publicar o comentar, y mide qué funciona post a post. Se puede automatizar como rutina diaria y semanal. Usar cuando alguien quiera construir presencia en LinkedIn, preparar contenido, buscar de qué hablar hoy, o entender por qué un post rindió como rindió.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.