Skip to content

Workflows agênticos — Maestria5 / 8

Custom Slash Commands as Team Workflows

A custom slash command is a reusable prompt you commit to the repo — so the whole team runs the same high-quality instruction instead of re-typing it.

Custom Slash Commands as Team Workflows

The best prompt your team ever wrote is worthless if everyone retypes a worse version of it. A custom slash command fixes that: it is a Markdown file whose body is a prompt, committed to the repo and run with /its-name.

The whole mechanism

Drop a file in .claude/commands/ (shared via git) or ~/.claude/commands/ (personal). The filename becomes the command:

<!-- .claude/commands/changelog.md -->
Summarize the staged git changes as a concise changelog entry,
grouped by Added / Changed / Fixed.

Now /changelog runs that prompt for everyone on the team.

Three features that make it powerful

Arguments$ARGUMENTS for everything, or $1, $2 for positional:

Fix issue #$1 with priority $2. Find the failing test, propose a patch,
explain the root cause.

Invoke: /fix-issue 4242 high.

Frontmatter configures the command:

---
description: Conventional-commit message from staged changes
argument-hint: [scope]
allowed-tools: Bash(git diff:*), Bash(git log:*)
model: claude-haiku-4-5
---

Inline context pulls live data in: prefix a line with ! to run a shell command and embed its output, or reference a file with @src/app.ts.

Why this is a team multiplier

Commands package intent; hooks enforce rules; MCP provides tools. The last building block is delegation — handing focused work to subagents.

Compartilhar este artigo

#Productivity #DevTools #AgenticAI

LinkedInX / TwitterBlueskyThreadsRedditHacker NewsWhatsAppE-mail

Série — Workflows agênticos — Maestria

  1. Parte 01A Mentalidade do Fluxo de Trabalho AgenticA maioria dos desenvolvedores trata IA como um autocomplete mais inteligente. Os que saem na frente a tratam como um colega de trabalho que age. Aqui está o modelo mental.
  2. Parte 02MCP Servers 101 — Dê Ferramentas Reais ao Seu AgentO Model Context Protocol é como seu agent para de adivinhar e começa a consultar seu banco de dados, seus issues, seu navegador. Aqui está o modelo mental e a primeira conexão.
  3. Parte 03Build Your First Custom MCP ServerOff-the-shelf servers cover GitHub and Postgres. The high-leverage one is the server only you can write — the bridge to your own system.
  4. Parte 04Hooks — Make the Agent Obey Your RulesA prompt asks the model to remember. A hook makes it happen — deterministically, every time, outside the model's control.
  5. Parte 05Custom Slash Commands as Team Workflowsvocê está aquiA custom slash command is a reusable prompt you commit to the repo — so the whole team runs the same high-quality instruction instead of re-typing it.
  6. Parte 06Subagentes — Delegando Trabalho que EscalaUm contexto gigante fica lento e vago. Subagentes deixam o agente principal delegar trabalho focado para especialistas com seu próprio contexto e ferramentas — e executá-los em paralelo.
  7. Parte 07The Daily-Driver Setup — Settings, Permissions, Status LineA diferença entre lutar contra o agent e fluir com ele é vinte minutos de configuração que você faz uma vez. Aqui está o setup.
  8. Parte 08Multi-Tool Pipelines — Ticket to Reviewed BranchThe payoff: chain MCP, hooks, commands and subagents into one flow that takes a ticket to a reviewed branch — with you in the loop only where it counts.

Continue aprendendo

PDF — vitalício

Comandos de barra

Todos os comandos de barra integrados, e como criar os seus.

Ver o PDF →

Curso

O curso Claude Mastery

12 módulos · 5 idiomas · certificado · teste 3 dias grátis.

Ver planos →
LinkedInX / TwitterBlueskyThreads