start-vibing 2.0.7 → 2.0.8

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "start-vibing",
-  "version": "2.0.7",
+  "version": "2.0.8",
   "description": "Setup Claude Code agents, skills, and hooks in your project. Smart copy that preserves your custom domains and configurations.",
   "type": "module",
   "bin": {

package/template/.claude/CLAUDE.md

@@ -17,6 +17,29 @@ This file provides context for all agents. For user-facing rules, see `/CLAUDE.m
 
 ---
 
+## MCP Servers (MANDATORY FOR AGENTS)
+
+All agents MUST use these MCP servers when applicable:
+
+| Server              | Purpose                                   | Use In Agents                            |
+| ------------------- | ----------------------------------------- | ---------------------------------------- |
+| `context7`          | Up-to-date library documentation          | research, analyzer, tester               |
+| `sequential-thinking` | Complex problem-solving                 | orchestrator, analyzer, final-validator  |
+| `memory`            | Persistent knowledge graph                | domain-updater, commit-manager           |
+| `playwright`        | Browser automation and E2E testing        | tester, ui-ux-reviewer                   |
+| `nextjs-devtools`   | Next.js specific development tools        | analyzer (Next.js projects)              |
+| `mongodb`           | MongoDB database operations               | tester, security-auditor                 |
+
+### Agent MCP Usage Rules
+
+- **research agent**: MUST use `context7` for library docs before recommending patterns
+- **tester agent**: MUST use `playwright` for E2E tests
+- **analyzer agent**: SHOULD use `context7` to verify current API patterns
+- **domain-updater**: SHOULD use `memory` to persist patterns across sessions
+- **ui-ux-reviewer**: MUST use `playwright` to verify UI implementations
+
+---
+
 ## Configuration Files
 
 Project-specific settings are in `.claude/config/`:

package/template/CLAUDE.md

@@ -4,16 +4,47 @@
 
 ---
 
-## Workflow
+## Workflow (STRICT FLOW)
 
 ```
-1. READ AGENTS    →  .claude/agents/*.md
-2. START TASK     →  python .claude/hooks/workflow-manager.py start-task --type feature --description "..."
-3. IMPLEMENT      →  Follow project rules
-4. QUALITY CHECK  →  Run all quality gates
-5. COMMIT         →  Conventional commits
+0. INIT TASK      →  Run commit-manager to REGISTER task start
+1. TODO LIST      →  ALWAYS create detailed todo list from prompt
+2. RESEARCH       →  Run research agent for NEW features (MANDATORY)
+3. AUDIT          →  Check docs from last audit OR run fresh audit
+4. CREATE BRANCH  →  feature/ | fix/ | refactor/ | test/
+5. IMPLEMENT      →  Follow project rules + analyzer approval
+6. DOCUMENT       →  Document ALL modified files (MANDATORY)
+7. QUALITY GATES  →  Run typecheck && lint && test (Husky enforced)
+8. FINISH         →  commit-manager creates PR with task summary
 ```
 
+> **CRITICAL:** Research agent is MANDATORY for any new feature or complex bug fix.
+> **CRITICAL:** All documentation must reference commit hashes for audit trail.
+> **CRITICAL:** Stop hook BLOCKS task completion if files not documented.
+> **FALLBACK:** If Task tool fails to invoke an agent, READ the agent file directly from `.claude/agents/[agent-name].md` and follow its instructions manually.
+
+---
+
+## MCP Servers (MANDATORY)
+
+Configure these in `.mcp.json` at project root:
+
+| Server              | Purpose                                   | When to Use                              |
+| ------------------- | ----------------------------------------- | ---------------------------------------- |
+| `context7`          | Up-to-date library documentation          | Before implementing with ANY library     |
+| `sequential-thinking` | Complex problem-solving                 | Multi-step tasks, planning               |
+| `memory`            | Persistent knowledge graph                | Store/recall project patterns            |
+| `playwright`        | Browser automation and E2E testing        | UI testing, page verification            |
+| `nextjs-devtools`   | Next.js specific development tools        | Next.js projects only                    |
+| `mongodb`           | MongoDB database operations               | Database queries, schema inspection      |
+
+### Usage Rules
+
+- **ALWAYS** use `context7` before implementing features with external libraries
+- **ALWAYS** use `playwright` for E2E tests instead of manual browser testing
+- **ALWAYS** check `memory` for existing patterns before creating new ones
+- **NEVER** hardcode library APIs - query `context7` for current docs
+
 ---
 
 ## Agent System
@@ -36,6 +67,18 @@ This project uses 11 specialized AI agents:
 
 ---
 
+## Stop Hook Enforcement
+
+The Stop hook runs when Claude is about to finish and BLOCKS if:
+
+| Violation | Action Required |
+|-----------|-----------------|
+| On main branch with source changes | Create feature/fix branch first |
+| Source files not documented | Run documenter agent |
+| Documenter not executed | Invoke documenter via Task tool |
+
+---
+
 ## Configuration
 
 Edit these files in `.claude/config/` for your project:
@@ -46,39 +89,17 @@ Edit these files in `.claude/config/` for your project:
 | `quality-gates.json`  | Quality check commands     |
 | `testing-config.json` | Test framework config      |
 | `security-rules.json` | Security audit rules       |
-
----
-
-## Quick Start
-
-```bash
-# 1. Start a task
-python .claude/hooks/workflow-manager.py start-task --type feature --description "Add user authentication"
-
-# 2. Approve files for modification
-python .claude/hooks/workflow-manager.py approve-files --files "src/auth/*.ts"
-
-# 3. Mark agents as executed
-python .claude/hooks/workflow-manager.py agent-executed --agent analyzer --result approved
-
-# 4. Run quality gates
-python .claude/hooks/workflow-manager.py quality-gate --gate typecheck --passed true
-
-# 5. Final validation
-python .claude/hooks/workflow-manager.py final-validation --result approved --ready-to-commit true
-
-# 6. Complete task
-python .claude/hooks/workflow-manager.py complete-task --commit-hash abc123
-```
+| `domain-mapping.json` | File patterns to domains   |
 
 ---
 
 ## Critical Rules
 
-1. **ALWAYS start task first** - Hooks block without active task
-2. **ALWAYS approve files** - Pre-tool-use hook blocks unapproved edits
+1. **ALWAYS use MCP servers** - Context7 for docs, Playwright for E2E
+2. **ALWAYS start with todo list** - Track progress with TodoWrite tool
 3. **NEVER skip agents** - Stop hook blocks incomplete workflows
 4. **ALWAYS use conventional commits** - feat:, fix:, refactor:, docs:
+5. **NEVER commit to main** - Create feature/fix branches
 
 ---