com um clique
Claude-Workflow
Claude-Workflow contém 11 skills coletadas de Mike-205, com cobertura ocupacional por repositório e páginas de detalhe dentro do site.
Skills neste repositório
Use this skill at the very start of a new project when the user wants to explore or explain their idea. Trigger when the user says things like "I have an idea", "I want to build something", "let me explain my project", "I want to create an app/website/tool", or begins describing a problem they want to solve. This is the non-technical first pass — focused on the problem, the people, and the desired outcome. Do NOT use for existing projects or when the user is already deep into technical decisions.
Use this skill after the non-technical discovery is complete and the user wants to define the technical design of the project. Trigger when the user says "let's talk tech", "what should the stack be", "let's design the architecture", "I want to plan the technical side", or after confirming DISCOVERY.md is accurate. Requires DISCOVERY.md to exist. Do NOT use before non-technical discovery is complete.
Use this skill to stress-test a plan, design, or architecture before building. Trigger when the user says "pressure test this", "grill me on this", "poke holes in my plan", "challenge my design", or after ARCHITECTURE.md is confirmed and the user says they're ready to proceed. Can also be triggered mid-project when a significant decision needs stress-testing. Reads DISCOVERY.md and ARCHITECTURE.md if they exist.
Use this skill to scaffold the initial project structure and template files after the architecture is designed and pressure-tested. Trigger when the user says "scaffold the project", "create the project structure", "set up the folders", or "let's start the project setup". Requires ARCHITECTURE.md to exist. Do NOT start writing real application logic — only structure and template files.
Use this skill to generate the project's own CLAUDE.md file after the project structure and git setup are complete. Trigger when the user says "create the CLAUDE.md", "generate claude context", or "set up claude for this project". Requires DISCOVERY.md, ARCHITECTURE.md, MILESTONES.md to exist. This generates the project CLAUDE.md — not the workflow one.
Use this skill to initialize git, connect the project to GitHub, and define the branch structure and milestone plan. Trigger when the user says "set up git", "connect to GitHub", "initialize git", or "set up version control". Should be run after project-structure is complete. Requires the project directory to exist and have a working structure.
Use this skill whenever committing code, staging changes, writing a commit message, or when the user says "commit this", "make a commit", "what should the commit message be", or "stage and commit". Applies to every commit throughout the project lifecycle. Do NOT use for PR descriptions or changelogs — those are separate.
Use this skill when starting work on a new milestone branch, or when the user says "let's start coding", "work on this branch", "build the next milestone", "continue the project", or "what do we build next". Reads CLAUDE.md and MILESTONES.md to understand the current branch goal. Do NOT start coding without confirming which branch we're on.
Use this skill to review code on the current branch before merging. Trigger when the user says "review this branch", "check my code", "review before merge", "code review", or after a milestone is complete. Reviews against the project's own architecture rules and conventions. Do NOT auto-approve — surface real issues.
Use this skill to write and run tests for the current branch. Trigger when the user says "write tests", "add tests", "test this", "run the tests", or after code review passes on a milestone. Reads ARCHITECTURE.md to use the correct testing framework. Tests are written for the current branch's code only.
Use this skill when the project (or a milestone) is complete and needs to be handed off to a client, another developer, or another agent. Trigger when the user says "prepare handoff", "write a handoff doc", "deliver to client", "summarize what was built", or "another agent will continue this". Also useful at end of a long session to compact context for the next session.