# Claude Code Configuration for {{PROJECT_NAME}}

## 📋 Quick Start

**IMPORTANT:** Read `AGENT.md` first for complete project context.

This file contains Claude Code-specific preferences and workflows.

---

## 🎯 Project Summary

**Name:** {{PROJECT_NAME}}

**Description:** {{PROJECT_DESCRIPTION}}

**Stack:** {{FRAMEWORK}} + {{LANGUAGE}} + {{DATABASE}}

---

## 📚 Documentation Architecture

**Always reference these files in order:**

1. `AGENT.md` - Universal AI configuration (READ FIRST)
2. `ai-instructions.md` - Tech stack, rules, workflow
3. `project-brief.md` - Business context
4. `README.md` - Developer onboarding & commands
5. `docs/architecture.md` - System architecture
6. `docs/data-model.md` - Database design
7. `docs/code-standards.md` - Code quality standards
8. `docs/testing.md` - Testing strategy
9. `docs/operations.md` - Deployment & operational procedures
10. `docs/business-flows.md` - Business process flows and diagrams
11. `docs/api.md` - API reference and conventions
12. `docs/contributing.md` - Contribution workflow
13. `specs/security.md` - Security policies
14. `specs/configuration.md` - Environment config

---

## ❌ CRITICAL RULES - NEVER

{{NEVER_RULES_SUMMARY}}

---

## ✅ CRITICAL RULES - ALWAYS

{{ALWAYS_RULES_SUMMARY}}

---

## 🏗️ Architecture Pattern

**Pattern:** {{ARCHITECTURE_PATTERN}}

**File Organization:** {{FILE_ORGANIZATION}}

---

## 🧪 Testing Requirements

- Minimum {{MIN_COVERAGE}}% coverage
- Write tests for all new features
- Run tests before committing: `{{TEST_COMMAND}}`

---

## 📝 Commit Format

{{COMMIT_FORMAT_EXAMPLE}}

---

## 🚀 Claude Code Workflow

### Workflow Modes
1. **🚀 Forced Agent Mode:** When running any command starting with `/flow-`, **IMMEDIATELY enter AGENT/EXECUTION MODE**. Do not wait for a general plan approval.
2. **Plan Mode:** Use plan mode for complex *custom* features that are not part of the standard `ai-flow` commands.
3. Break down custom features into small, testable chunks.
4. Get user approval before implementation for custom features.

### Implementation
1. Follow architecture patterns strictly
2. Write tests alongside code
3. Update documentation if behavior changes

---

## 🔒 Security Checklist

- [ ] Input validation with {{VALIDATION_LIBRARY}}
- [ ] Authentication/authorization checked
- [ ] No hardcoded secrets
- [ ] No sensitive data in logs
- [ ] Rate limiting for public endpoints

---

## 💡 Claude-Specific Tips

- Use `/flow-build` command for starting documentation generation
- Use plan mode for architectural decisions
- Reference `AGENT.md` for any clarifications
- Ask user before making breaking changes

---

**For full context, always read `AGENT.md` and `ai-instructions.md` before starting work.**

**Generated by:** AI Flow v1.0.0