Use when the user wants to build a feature, implement functionality, fix a bug, create something new, or do any development work. Handles the complete workflow from brainstorming through implementation with research, design, planning, and quality gates. Invoke this skill for requests like "I want to add...", "Build me a...", "Create a...", "Implement...", "Fix...".
/plugin marketplace add astrosteveo/claude-code-plugins/plugin install astrosteveo-explore-plan-implement-commit-plugins-explore-plan-implement-commit@astrosteveo/claude-code-pluginsThis skill inherits all available tools. When active, it can use any tool Claude has access to.
The main entry point for the plugin. Handles intent detection, phase progression, and orchestrates the three subagents.
| User says | Plugin understands |
|---|---|
| "I wanna make a..." | New feature → maybe brainstorm first, then Explore |
| "This isn't working..." | Bug fix → start at Explore (skip brainstorm) |
| "Resume 001" | Resume → check artifacts, pick up where left off |
| "Add dark mode" | Clear request → start at Explore |
If request is vague or creative:
| Step | What | Output | Commit |
|---|---|---|---|
| Map codebase | Discoverer documents existing code | codebase.md | ✓ |
| Research | Discoverer fetches best practices | research.md | ✓ |
| Step | What | Output | Commit |
|---|---|---|---|
| Design | Architecture decisions, Mermaid diagrams | design.md | ✓ |
| Tasks | Break into atomic, ordered tasks | plan.md | ✓ |
| Step | What | Output | Commit |
|---|---|---|---|
| Per task | Coder implements, Reviewer validates | code + tests | ✓ each |
| Update | Mark task complete with hash | plan.md | — |
| Step | What | Output | Commit |
|---|---|---|---|
| Summarize | Document what was built | summary | ✓ |
| PR | Create PR if on feature branch | — | — |
.workflow/
├── 001-user-authentication/
│ ├── codebase.md
│ ├── research.md
│ ├── design.md
│ ├── plan.md
│ └── issues.md
├── 002-dark-mode/
│ └── ...
└── backlog.md
Check which artifacts exist:
plan.md with unchecked tasks? → Resume at Codedesign.md but no plan.md? → Resume at Plan (tasks)codebase.md but no research.md? → Resume at Explore (research)Task 1 → Reviewer ─┐
Task 2 → Reviewer ─┼→ User checkpoint
Task 3 → Reviewer ─┘
...
User reviews every 2-3 tasks.
Task 1 → Reviewer → Task 2 → Reviewer → ... → Done → User review
No pauses until complete.
Both modes: Reviewer runs after every task. Quality gates are non-negotiable.
When creating design.md:
codebase.md and research.md| Type | Use for |
|---|---|
| Flowchart | Process flows, decision logic |
| Sequence | Component interactions |
| Class | Data models, interfaces |
| ER | Database schema |
| State | Lifecycle, state machines |
When creating plan.md:
- [ ] Task 1: Set up auth routes
- Create `src/routes/auth.ts`
- Add login/logout endpoints
- Expected: routes respond to requests
- [ ] Task 2: Add validation
- Create `src/validation/auth.ts`
- Validate email format, password strength
- Expected: invalid inputs rejected
Update as tasks complete:
- [x] Task 1: Set up auth routes *(committed: abc1234)*
- [x] Task 2: Add validation *(committed: def5678)*
- [ ] Task 3: Create login form
<stage>: <what was done>
explore: document codebase for user-auth
explore: research JWT best practices
plan: design authentication architecture
plan: create implementation tasks
code: implement login route
code: add session middleware
commit: finalize user-auth feature
| Subagent | When | For |
|---|---|---|
| Discoverer | Explore phase | Codebase mapping, research |
| Coder | Code phase | Task implementation |
| Reviewer | After each task | Quality validation |
The workflow skill coordinates; subagents execute.
This skill should be used when the user asks to "create an agent", "add an agent", "write a subagent", "agent frontmatter", "when to use description", "agent examples", "agent tools", "agent colors", "autonomous agent", or needs guidance on agent structure, system prompts, triggering conditions, or agent development best practices for Claude Code plugins.
This skill should be used when the user asks to "create a slash command", "add a command", "write a custom command", "define command arguments", "use command frontmatter", "organize commands", "create command with file references", "interactive command", "use AskUserQuestion in command", or needs guidance on slash command structure, YAML frontmatter fields, dynamic arguments, bash execution in commands, user interaction patterns, or command development best practices for Claude Code.
This skill should be used when the user asks to "create a hook", "add a PreToolUse/PostToolUse/Stop hook", "validate tool use", "implement prompt-based hooks", "use ${CLAUDE_PLUGIN_ROOT}", "set up event-driven automation", "block dangerous commands", or mentions hook events (PreToolUse, PostToolUse, Stop, SubagentStop, SessionStart, SessionEnd, UserPromptSubmit, PreCompact, Notification). Provides comprehensive guidance for creating and implementing Claude Code plugin hooks with focus on advanced prompt-based hooks API.