ai-flow-dev 2.2.0 → 2.2.4

package/prompts/shared/smart-skip-preflight.md

@@ -0,0 +1,214 @@
+# Smart Skip Pre-Flight Check (Shared Template)
+
+## 🔍 Pre-Flight Check Logic
+
+**This section is automatically included in Phases 1-7 to enable smart skip functionality.**
+
+### How It Works
+
+1. **Read audit data** from Phase 0: `.ai-flow/cache/audit-data.json`
+2. **Check consistency score** for the current phase
+3. **Execute appropriate scenario** based on score
+
+---
+
+## Execution Flow
+
+```javascript
+// Read audit data from Phase 0
+const auditData = readJSON('.ai-flow/cache/audit-data.json');
+const phaseData = auditData?.phases?.[currentPhase];
+
+if (phaseData?.exists) {
+  // Documentation exists, check consistency
+  if (phaseData.consistencyScore >= 95) {
+    executeScenarioA(); // SKIP
+  } else if (phaseData.consistencyScore >= 80) {
+    executeScenarioB(); // HYBRID
+  } else {
+    executeScenarioC(); // FULL
+  }
+} else {
+  // No existing docs, proceed with full phase
+  executeScenarioC();
+}
+```
+
+---
+
+## Scenario A: High Confidence (≥95%) - SKIP
+
+**Display format:**
+
+```
+---
+✅ [PHASE NAME] ALREADY DOCUMENTED
+
+File: [target-file]
+Consistency: [score]%
+Status: Complete documentation
+
+Documented Information:
+✅ [Key item 1]
+✅ [Key item 2]
+✅ [Key item 3]
+...
+
+---
+
+Use existing documentation? (Y/n) ⭐
+> _
+```
+
+**User Actions:**
+- **Y**: Skip to next phase
+- **n**: Proceed with full phase questionnaire
+
+---
+
+## Scenario B: Medium Confidence (80-94%) - HYBRID
+
+**Display format:**
+
+```
+---
+📚 [PHASE NAME] PARTIALLY DOCUMENTED
+
+File: [target-file]
+Consistency: [score]%
+Status: Most information present, some gaps
+
+Documented:
+✅ [Item 1]
+✅ [Item 2]
+
+Missing:
+❌ [Gap 1]
+❌ [Gap 2]
+
+---
+
+Options:
+A) Answer only [N] missing questions ⭐
+B) Re-answer everything (full phase)
+C) Skip entirely (use docs as-is)
+
+> _
+```
+
+**User Actions:**
+- **A**: Ask only missing questions, merge with existing docs
+- **B**: Full phase questionnaire
+- **C**: Skip with warning
+
+---
+
+## Scenario C: Low Confidence (<80%) or No Docs - FULL PHASE
+
+**Display format:**
+
+```
+---
+⚠️ [PHASE NAME] NEEDS DOCUMENTATION
+
+Reason: [target-file] is outdated or missing
+Consistency: [score]% (or N/A if file doesn't exist)
+
+Issues detected:
+❌ [Issue 1]
+❌ [Issue 2]
+
+Recommendation: Complete full phase questionnaire
+
+Proceeding with full phase...
+[Continue to first question]
+```
+
+---
+
+## Phase-Specific Configuration
+
+Each phase references this template with specific parameters:
+
+### Phase 1: Discovery & Business
+- **Target File**: `project-brief.md`
+- **Phase Name**: "BUSINESS CONTEXT"
+- **Key Items**: Project name, description, users, objectives, system type, features, scope, constraints, metrics, business flows
+- **Typical Gaps**: Business objectives, success metrics, constraints
+
+### Phase 2: Data Architecture
+- **Target File**: `docs/data-model.md`
+- **Phase Name**: "DATA ARCHITECTURE"
+- **Key Items**: Entities, relationships, data patterns, indexes
+- **Typical Gaps**: Missing entities, undocumented relationships, missing fields
+
+### Phase 3: System Architecture
+- **Target File**: `docs/architecture.md`
+- **Phase Name**: "SYSTEM ARCHITECTURE"
+- **Key Items**: Framework, architecture pattern, API style, database, caching, background jobs, integrations
+- **Typical Gaps**: API versioning, rate limiting, caching strategy
+
+### Phase 4: Security & Authentication
+- **Target File**: `specs/security.md`
+- **Phase Name**: "SECURITY & AUTHENTICATION"
+- **Key Items**: Auth strategy, encryption, security patterns, compliance
+- **Typical Gaps**: Compliance requirements, audit logging, security policies
+
+### Phase 5: Code Standards
+- **Target File**: `docs/code-standards.md`
+- **Phase Name**: "CODE STANDARDS"
+- **Key Items**: Linters, formatters, naming conventions, code review process
+- **Typical Gaps**: Team-specific conventions, code review workflow
+
+### Phase 6: Testing Strategy
+- **Target File**: `docs/testing.md`
+- **Phase Name**: "TESTING STRATEGY"
+- **Key Items**: Test framework, coverage targets, test types, CI/CD integration
+- **Typical Gaps**: E2E strategy, load testing, performance testing
+
+### Phase 7: Operations & Deployment
+- **Target File**: `docs/deployment.md`
+- **Phase Name**: "OPERATIONS & DEPLOYMENT"
+- **Key Items**: CI/CD pipeline, deployment platform, monitoring, logging
+- **Typical Gaps**: Incident runbooks, disaster recovery, scaling strategy
+
+---
+
+## Usage in Phase Prompts
+
+**In each phase prompt (1-7), add this reference:**
+
+```markdown
+## 🔍 Pre-Flight Check (Smart Skip Logic)
+
+> 📎 **Reference:** See [prompts/shared/smart-skip-preflight.md](../shared/smart-skip-preflight.md) for the complete smart skip logic.
+
+**Execute Pre-Flight Check for this phase:**
+
+- **Phase**: [Phase Number]
+- **Target File**: [file path]
+- **Phase Name**: [display name]
+
+[Proceed with appropriate scenario based on audit data]
+
+---
+
+## Phase [N] Questions (Full Mode)
+
+[Continue with phase-specific questions...]
+```
+
+---
+
+## Benefits of This Approach
+
+✅ **DRY Principle**: Single source of truth for skip logic
+✅ **Easy Maintenance**: Update once, applies to all phases
+✅ **Consistency**: Same UX across all phases
+✅ **Reduced File Size**: Each phase prompt is ~100 lines shorter
+✅ **Clear Separation**: Logic vs content
+
+---
+
+**Last Updated:** 2025-12-22
+**Version:** 1.0

package/templates/AGENT.template.md

@@ -141,7 +141,12 @@ All documentation is structured to guide AI assistants in understanding the proj
 
 When working on this project:
 
-1. **Before starting ANY task:**
+1. **🚀 Proactive Execution (MANDATORY):**
+   - When any command starting with `/flow-` is detected (e.g., `/flow-build`, `/flow-work`, `/flow-check`), **IMMEDIATELY enter AGENT/EXECUTION MODE**.
+   - **Do NOT ask for permission** to read files, write documents, or execute shell commands that are part of the interactive flow.
+   - These flows already include their own internal planning and confirmation steps. Proactivity is key to efficiency.
+
+2. **Before starting ANY task:**
    - Read `ai-instructions.md` for project-wide rules
    - Check relevant documentation for the area you're working on
    - Understand the business context from `project-brief.md`
@@ -336,7 +341,12 @@ All documentation is structured to guide AI assistants in understanding the proj
 
 When working on this project:
 
-1. **Before starting ANY task:**
+1. **🚀 Proactive Execution (MANDATORY):**
+   - When any command starting with `/flow-` is detected (e.g., `/flow-build`, `/flow-work`, `/flow-check`), **IMMEDIATELY enter AGENT/EXECUTION MODE**.
+   - **Do NOT ask for permission** to read files, write documents, or execute shell commands that are part of the interactive flow.
+   - These flows already include their own internal planning and confirmation steps. Proactivity is key to efficiency.
+
+2. **Before starting ANY task:**
    - Read `ai-instructions.md` for project-wide rules
    - Check relevant documentation for the area you're working on
    - Understand the business context from `project-brief.md`
@@ -412,4 +422,4 @@ See `README.md` for installation and setup instructions.
 
 **Last Updated:** {{GENERATION_DATE}}
 
-**Generated by:** AI Flow v2.1.9
+**Generated by:** AI Flow v2.2.4

package/templates/backend/.clauderules.template

@@ -75,10 +75,11 @@ This file contains Claude Code-specific preferences and workflows.
 
 ## 🚀 Claude Code Workflow
 
-### Plan Mode
-1. Always use plan mode for complex features
-2. Break down into small, testable chunks
-3. Get user approval before implementation
+### 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

package/templates/backend/.cursorrules.template

@@ -92,7 +92,7 @@ Read these files in order:
 
 - Use `@AGENT.md` to reference project context
 - Use `@ai-instructions.md` for coding rules
-- Run `/flow-build` to start documentation generation
+- **🚀 Proactive Execution:** When running `/flow-build`, `/flow-work`, or `/flow-check`, **ACT IMMEDIATELY**. These flows have their own interactive confirmation steps. Do not wait for a general plan approval.
 - Check `docs/` folder for detailed specifications
 
 ---