@iservu-inc/adf-cli 0.9.0 → 0.10.0

package/CHANGELOG.md

@@ -5,8 +5,460 @@ All notable changes to `@iservu-inc/adf-cli` will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [0.10.0] - 2025-10-27
+
+### 🧠 Pattern Decay Algorithm - Phase 6
+
+**MAJOR FEATURE:** Intelligent pattern decay system that keeps learning data fresh and relevant over time.
+
+#### What's New
+
+**Pattern Decay System:**
+- **Time-Based Exponential Decay** - Patterns lose confidence over time when inactive
+- **Confidence-Based Decay Rates** - Stronger patterns fade slower than weaker ones
+- **Automatic Pattern Cleanup** - Removes stale patterns below threshold or inactive 6+ months
+- **Pattern Renewal System** - Reconfirmed patterns get confidence boost (+10 points)
+- **User-Approved Protection** - Manually approved patterns decay at half rate
+
+#### Technical Implementation
+
+**Decay Formula:**
+```javascript
+newConfidence = currentConfidence × (1 - decayRate)^monthsInactive
+```
+
+**Decay Rates by Confidence:**
+- High confidence (≥90%): 7.5% per month (decay slower)
+- Medium confidence (75-89%): 15% per month (normal decay)
+- Low confidence (<75%): 22.5% per month (decay faster)
+
+**Pattern Lifecycle:**
+- **Active** - Pattern confidence ≥40, seen within 6 months
+- **Renewed** - Pattern reconfirmed, gets +10 confidence boost
+- **Removed** - Confidence <40 OR inactive 6+ months
+
+#### New Components
+
+**DecayManager (lib/learning/decay-manager.js):**
+- `loadPatternsWithDecay()` - Automatic decay on pattern load
+- `checkForPatternRenewal()` - Match skip events to patterns and renew
+- `cleanupStalePatterns()` - Separate active from stale patterns
+- `triggerDecayCalculation()` - Manual decay execution
+- `getDecayStats()` - Analytics and reporting
+- Removal history tracking
+
+**Enhanced Pattern Detection (lib/learning/pattern-detector.js):**
+- `calculateDecay()` - Apply exponential decay to single pattern
+- `getDecayRate()` - Dynamic rate calculation based on confidence
+- `renewPattern()` - Boost pattern confidence on reconfirmation
+- `applyDecayToPatterns()` - Batch decay application
+- `cleanupStalePatterns()` - Identify and remove stale patterns
+- `enhancePatternsWithMetadata()` - Add decay timestamps
+
+**Storage Integration (lib/learning/storage.js):**
+- `getDecayConfig()` - Load decay configuration
+- `saveDecayConfig()` - Save decay configuration
+- Default decay config in `getLearningConfig()`
+
+#### Configuration
+
+**Default Decay Settings:**
+```json
+{
+  "enabled": true,
+  "baseDecayRate": 0.15,              // 15% per month
+  "minConfidenceThreshold": 50,       // Don't filter if < 50
+  "removeBelow": 40,                  // Auto-remove if < 40
+  "renewalBoost": 10,                 // +10 points on renewal
+  "protectApproved": true,            // User-approved decay at 0.5x
+  "highConfidenceThreshold": 90,      // ≥90 decay slower
+  "lowConfidenceThreshold": 75,       // <75 decay faster
+  "maxRenewalsPerDay": 1,             // Limit inflation
+  "maxInactiveMonths": 6              // Auto-remove after 6 months
+}
+```
+
+Access via `adf config` → Learning System (coming soon).
+
+#### Pattern Metadata
+
+**New Timestamp Fields:**
+- `createdAt` - When pattern was first detected
+- `lastSeen` - Last time pattern was observed (skip event matched)
+- `lastDecayCalculation` - Last time decay was applied
+- `initialConfidence` - Original confidence score (for tracking history)
+- `timesRenewed` - Count of how many times pattern was reconfirmed
+
+#### Use Cases
+
+**Scenario 1: Workflow Change**
+- User builds prototypes for 3 months → skips deployment questions (90% confidence)
+- User switches to production apps → needs deployment questions
+- Old pattern decays over time → eventually removed automatically
+- ✅ System adapts to workflow changes
+
+**Scenario 2: Pattern Reconfirmation**
+- Pattern confidence decayed to 60%
+- User skips same question again
+- Pattern renewed: confidence increases to 70% (+10 boost)
+- `lastSeen` updated, decay timer resets
+- ✅ Active patterns stay strong
+
+#### Testing
+
+**Test Coverage:**
+- 25+ unit tests for decay calculations (tests/pattern-decay.test.js)
+- 15+ integration tests for DecayManager (tests/decay-manager.test.js)
+- Edge case testing (zero confidence, very old patterns, oscillation)
+- All tests passing ✅
+
+**Test Scenarios:**
+- Exponential decay over time
+- Confidence-based decay rates
+- Pattern renewal with confidence boost
+- User-approved pattern protection
+- Automatic cleanup of stale patterns
+- Removal history tracking
+- Empty pattern list handling
+
+#### Files Added
+- `lib/learning/decay-manager.js` - Orchestration layer for decay operations
+- `tests/pattern-decay.test.js` - Unit tests for decay calculations
+- `tests/decay-manager.test.js` - Integration tests for DecayManager
+- `.project/docs/designs/PATTERN-DECAY-ALGORITHM.md` - Design specification
+
+#### Files Modified
+- `lib/learning/pattern-detector.js` - Added decay calculation methods
+- `lib/learning/storage.js` - Added decay config persistence
+- `package.json` - Version 0.9.1 → 0.10.0
+
+#### Why This Matters
+
+**Before:** Patterns remained at high confidence indefinitely, causing stale patterns to incorrectly filter questions even after user workflows changed.
+
+**After:** Patterns naturally decay when inactive, allowing the learning system to adapt to evolving user workflows while keeping frequently-used patterns strong through renewal.
+
+#### Performance
+
+- **Computational:** O(n) where n = number of patterns (~10-50)
+- **Execution Time:** <10ms for 50 patterns
+- **Storage Impact:** ~112 bytes per pattern for new fields
+- **Memory:** Negligible (5.5 KB for 50 patterns)
+
+#### Next Steps
+
+Integration into interview workflow will be completed in a future release to:
+- Apply decay automatically on interview start
+- Renew patterns when skip events match
+- Show decay statistics in learning dashboard
+- Add decay configuration UI in `adf config`
+
+---
+
+## [0.9.1] - 2025-10-05
+
+### 🐛 Critical Bug Fix
+
+**Fixed: Syntax Error in interviewer.js**
+- **Problem:** Syntax error in `interviewer.js` causing interview to fail on startup
+- **Root Cause:** Missing closing parenthesis in method call
+- **Solution:** Fixed syntax error at line 380
+- **Impact:** Interview now starts correctly without errors
+
+**Technical Details:**
+- Modified: `lib/frameworks/interviewer.js:380` - Fixed missing closing parenthesis
+- All 200 tests passing
+
+---
+
+## [0.9.0] - 2025-10-05
+
+### 🎛️ AI Analysis Settings & Performance Modes
+
+**MAJOR FEATURE:** Configurable AI analysis with three performance modes - giving users complete control over speed vs intelligence tradeoff.
+
+#### Three Performance Modes
+
+**Fast Mode:**
+- Zero AI delays (instant responses)
+- All AI features disabled
+- Best for: Quick workflows, prototyping, low-priority projects
+- Speed: ~0.5s per answer
+
+**Balanced Mode (Default):**
+- 2-3 seconds per answer
+- Quality Analysis enabled
+- Smart Filtering enabled
+- Pattern Detection enabled
+- Question Reordering disabled
+- Follow-up Questions disabled
+- Best for: Most projects
+- Optimal balance of speed and intelligence
+
+**Comprehensive Mode:**
+- 4-6 seconds per answer
+- All AI features enabled
+- Maximum intelligence and guidance
+- Best for: Complex projects, critical systems
+- Most thorough analysis
+
+#### Configurable AI Features
+
+Five granular AI features can be toggled individually:
+
+1. **AI-Powered Quality Analysis** (Medium Impact: 1-2s)
+   - Analyzes answer quality in real-time
+   - Provides 0-100 quality scores
+   - Suggests improvements for low-quality answers
+
+2. **Intelligent Question Reordering** (High Impact: 2-3s)
+   - Reorders questions based on extracted knowledge
+   - Prioritizes fundamental questions first
+   - Adapts interview flow dynamically
+
+3. **AI-Generated Follow-Up Questions** (Medium Impact: 1-2s when triggered)
+   - Creates context-specific follow-up questions
+   - Only triggers for low-quality answers (< 70 score)
+   - Helps gather missing information
+
+4. **Pattern Detection & Learning** (Low Impact: minimal)
+   - Learns from your skip behavior
+   - Generates learned rules over time
+   - Always lightweight, runs locally
+
+5. **Smart Question Filtering** (Low Impact: minimal)
+   - Analyzes project type and context
+   - Filters out irrelevant questions
+   - Saves time on specialized projects
+
+#### New Configuration Category
+
+**AI Analysis Settings** added to `adf config`:
+```
+⚙️  ADF Configuration
+
+? Select configuration category:
+  AI Provider Setup - ✓ Configured (OpenAI)
+  AI Analysis Settings - Balanced Mode ⚙️ NEW
+  IDE Deployment - ✓ Deployed (Windsurf, Cursor)
+  Learning System - ✓ Active (5 sessions, 12 patterns)
+```
+
+**Configuration Menu:**
+- Performance Mode selection (Fast/Balanced/Comprehensive)
+- Individual feature toggles with performance impact indicators
+- Feature descriptions explaining each capability
+- Settings saved to `.adf/analysis-config.json`
+
+#### User Experience
+
+**At Interview Start:**
+```
+🎛️  AI Analysis Mode: Balanced
+   (Configure via: adf config → AI Analysis Settings)
+```
+
+**During Configuration:**
+- Clear descriptions of each mode
+- Performance impact indicators (High/Medium/Low)
+- Recommendations for different project types
+- Instant preview of changes
+
+#### Benefits
+
+- **User Control:** Complete control over AI features and performance
+- **Flexibility:** Choose speed or intelligence based on project needs
+- **Transparency:** Clear impact indicators for each feature
+- **Efficiency:** Fast mode for quick workflows, Comprehensive for critical work
+- **Progressive Enhancement:** Works with or without AI provider configured
+
+#### Technical Implementation
+
+**New Files:**
+- `lib/analysis/analysis-config.js` - Configuration management and defaults
+- `.adf/analysis-config.json` - User's analysis settings (gitignored)
+
+**Modified Files:**
+- `lib/commands/config.js` - Added AI Analysis Settings category
+- `lib/frameworks/interviewer.js` - Load and display analysis mode
+- `lib/analysis/dynamic-pipeline.js` - Respect feature toggles
+- `lib/filters/question-filter.js` - Respect filtering toggle
+- `lib/learning/learning-manager.js` - Respect learning toggles
+
+**Configuration Schema:**
+```javascript
+{
+  mode: "balanced",  // fast | balanced | comprehensive
+  features: {
+    qualityAnalysis: true,
+    questionReordering: false,
+    followUpQuestions: false,
+    patternDetection: true,
+    smartFiltering: true
+  }
+}
+```
+
+#### Testing
+
+- ✅ All 200 tests passing
+- ✅ Backward compatible with existing sessions
+- ✅ Graceful degradation if config missing
+- ✅ Works with or without AI provider
+
+---
+
+## [0.8.0] - 2025-10-05
+
+### 💾 Resume from Exit & Graceful Quit
+
+**MAJOR FEATURE:** Resume interviews after exit and graceful Ctrl+C handling.
+
+#### Resume Capability
+
+**Exit Anytime:**
+- Type `exit` at any question to save progress and quit
+- Press `Ctrl+C` anywhere to trigger graceful exit
+- Progress automatically saved before exiting
+- Resume with `adf init` to continue from last question
+
+**Resume Flow:**
+```
+$ adf init
+
+📦 Resumable Session Detected
+
+Session: my-project-session
+Started: 2025-10-05 14:30
+Questions answered: 8 / 20
+Information richness: 65%
+
+? What would you like to do?
+  ❯ Resume Interview
+    Start New Session
+    Exit
+```
+
+**Smart Skip on Resume:**
+- Already-answered questions automatically skipped
+- Shows "✓ Already answered" with answer preview
+- Resumes exactly where you left off
+- No redundant questions
+
+#### Ctrl+C Graceful Quit
+
+**Global Handler:**
+- Catches `Ctrl+C` (SIGINT) anywhere in interview
+- Saves progress automatically before exit
+- Shows confirmation message
+- Clean exit (no error messages)
+
+**Works From:**
+- Question prompts
+- Menu selections
+- AI configuration
+- Anywhere during interview
+
+#### Benefits
+
+- **Flexibility:** Exit and resume anytime without losing work
+- **Safety:** Automatic progress saving on exit
+- **Convenience:** Both `exit` command and `Ctrl+C` supported
+- **Transparency:** Clear messaging about saved progress
+- **User Control:** Choose to resume or start new session
+
+#### Technical Implementation
+
+**Modified Files:**
+- `lib/frameworks/interviewer.js`:
+  - Added exit command detection
+  - Added Ctrl+C handler registration
+  - Skip already-answered questions on resume
+  - Show answer preview for skipped questions
+  - Clean exit messaging
+
+**Resume Detection:**
+- Checks `_progress.json` for existing session
+- Validates session state
+- Prompts user to resume or start new
+- Preserves all context (answers, quality metrics, blocks)
+
+#### Testing
+
+- ✅ All 200 tests passing
+- ✅ Exit at any question works
+- ✅ Ctrl+C handler works globally
+- ✅ Resume from exact position works
+- ✅ Already-answered questions skipped correctly
+
+---
+
+## [0.7.1] - 2025-10-05
+
+### 🔄 Reverted to Terminal Input
+
+**UX FIX:** Reverted external editor approach (v0.6.0-v0.7.0) back to simple terminal input.
+
+#### What Changed
+
+**Reverted:**
+- ❌ External editor (Notepad, nano, vi) - Bad UX, too disruptive
+- ❌ Multi-line input experiment - Added complexity without value
+- ✅ Back to simple terminal input with exit/skip shortcuts
+
+**Why We Reverted:**
+- External editor was jarring and disruptive to workflow
+- Opening Notepad for every answer broke flow state
+- Users prefer inline typing despite cursor lag
+- Cursor lag accepted as terminal limitation (not critical)
+
+**What We Kept:**
+- ✅ Exit shortcut functionality (type "exit" to quit)
+- ✅ Skip shortcut (type "skip" to skip questions)
+- ✅ Resume capability from v0.8.0
+
+#### Current Behavior
+
+**Answer Input:**
+- Type directly in terminal (single line)
+- Type "skip" to skip remaining questions in block
+- Type "exit" to save and quit (resume later)
+- Press Ctrl+C for graceful quit
+- Cursor lag accepted as terminal quirk
+
+#### Benefits
+
+- **Better UX:** Inline typing feels more natural
+- **No disruption:** No external editor popup
+- **Fast workflow:** Minimal friction
+- **User preference:** Community feedback validated inline approach
+
+#### Technical Details
+
+- Reverted: `lib/frameworks/interviewer.js` - Back to `type: 'input'`
+- Removed: Editor workflow code and instructions
+- Kept: Exit and skip functionality
+- All 200 tests passing
+
+---
+
+## [0.7.0] - 2025-10-05
+
+### ⚠️ Note: This version was reverted in v0.7.1
+
+**Exit Shortcut Feature:**
+- Added `exit` command to quit interview anytime
+- Type "exit" at any question to save progress and quit
+- Resume with `adf init` to continue
+
+**Note:** External editor approach from v0.6.0 was continued in this version but later reverted in v0.7.1 based on user feedback.
+
+---
+
 ## [0.6.0] - 2025-10-05
 
+### ⚠️ Note: This version was reverted in v0.7.1
+
 ### 🎯 Fixed Cursor Positioning with Editor Input
 
 **FIX:** Switched from inline input to editor-based input to fix cursor lag on multi-line answers.

package/README.md

@@ -62,7 +62,7 @@ adf init --rapid --tool cursor
 
 ### Configure ADF Settings
 
-Configure ADF settings like AI provider, learning system, and more:
+Configure ADF settings like AI provider, AI analysis settings, learning system, and more:
 
 ```bash
 adf config
@@ -70,14 +70,17 @@ adf config
 
 This command provides an interactive menu with:
 - **AI Provider Setup** - Configure Anthropic, OpenAI, Google Gemini, or OpenRouter
+- **AI Analysis Settings** - Configure performance modes and AI features
+- **IDE Deployment** - Deploy to multiple IDEs
 - **Learning System** - Manage interview learning data and preferences
 - Status indicators showing what's configured (green ✓), disabled (yellow ○), or has data
-- Future configuration options (coming soon)
 
 You can run `adf config` anytime to:
 - Configure AI for the first time
 - Switch AI providers or models
 - Update your API keys
+- Adjust AI analysis performance and features
+- Deploy to development tools
 - Review learning patterns and manage learned preferences
 
 ### Deploy to Development Tool
@@ -203,6 +206,78 @@ Select "AI Provider Setup" and you'll see:
 
 Your previous API keys remain saved in `.adf/.env` for easy switching between providers.
 
+## AI Analysis Settings
+
+ADF CLI provides three performance modes and five configurable AI features, giving you complete control over the speed vs intelligence tradeoff during interviews.
+
+### Performance Modes
+
+Configure via `adf config` → AI Analysis Settings:
+
+**Fast Mode:**
+- Zero AI delays (~0.5s per answer)
+- All AI features disabled
+- Best for: Quick workflows, prototyping, low-priority projects
+
+**Balanced Mode (Default):**
+- 2-3 seconds per answer
+- Quality Analysis, Smart Filtering, and Pattern Detection enabled
+- Question Reordering and Follow-up Questions disabled
+- Best for: Most projects - optimal balance of speed and intelligence
+
+**Comprehensive Mode:**
+- 4-6 seconds per answer
+- All AI features enabled
+- Maximum intelligence and guidance
+- Best for: Complex projects, critical systems, maximum thoroughness
+
+### Configurable AI Features
+
+Five AI features can be toggled individually:
+
+1. **AI-Powered Quality Analysis** (Medium Impact: 1-2s)
+   - Real-time answer quality scoring (0-100)
+   - Improvement suggestions for low-quality answers
+   - Helps ensure comprehensive requirements gathering
+
+2. **Intelligent Question Reordering** (High Impact: 2-3s)
+   - Dynamically reorders questions based on extracted knowledge
+   - Prioritizes fundamental questions first
+   - Adapts interview flow to your answers
+
+3. **AI-Generated Follow-Up Questions** (Medium Impact: 1-2s when triggered)
+   - Context-specific follow-ups for incomplete answers
+   - Only triggers for low-quality answers (< 70 score)
+   - Helps gather missing information intelligently
+
+4. **Pattern Detection & Learning** (Low Impact: minimal)
+   - Learns from your skip behavior over time
+   - Generates learned rules from patterns
+   - Lightweight, runs locally
+
+5. **Smart Question Filtering** (Low Impact: minimal)
+   - Analyzes project type and context
+   - Filters out irrelevant questions automatically
+   - Saves time on specialized projects (CLI tools, APIs, etc.)
+
+### Configuration
+
+Access AI Analysis Settings:
+
+```bash
+adf config
+# Select "AI Analysis Settings"
+```
+
+Settings are saved to `.adf/analysis-config.json` and apply to all future interviews.
+
+**Performance Mode Display:**
+At interview start, you'll see:
+```
+🎛️  AI Analysis Mode: Balanced
+   (Configure via: adf config → AI Analysis Settings)
+```
+
 ## Learning System
 
 ADF CLI includes an intelligent Learning System that improves your interview experience by learning from your behavior across sessions.
@@ -414,27 +489,41 @@ When we release updates to the framework:
 
 See [CHANGELOG.md](./CHANGELOG.md) for detailed version history.
 
-**Latest:** v0.4.12
-- **Learning System (Phase 4.2)** - Learns from your skip behavior to improve filtering
-  - Automatic pattern detection across interview sessions
-  - Learned rule generation from high-confidence patterns
-  - Adaptive filtering with user approval
-  - Learning management CLI via `adf config`
-  - Privacy-first local storage in `.adf/learning/`
-- **Smart Question Filtering (Phase 4.1)** - Context-aware question filtering
-  - Project analysis (type, frameworks, languages)
-  - Intelligent question relevance scoring
-  - Time estimation for skipped questions
-  - User control with confidence thresholds
-- Enhanced config command with Learning System integration
-
-**Previous:** v0.3.6
-- AI-guided requirements gathering with multi-provider support
-- Real-time answer quality analysis and intelligent follow-ups
-- Secure API key persistence in `.adf/.env`
-- Dynamic model fetching with autocomplete selection
-- Enhanced UX with visible skip instructions
-- Automatic deployment to your preferred IDE
+**Latest:** v0.10.0 (2025-10-27)
+- **Pattern Decay Algorithm (v0.10.0)** - Phase 6: Advanced Learning Features
+  - Time-based exponential decay for inactive patterns
+  - Confidence-based decay rates (high/medium/low)
+  - Automatic pattern cleanup and renewal system
+  - 40+ comprehensive tests for decay functionality
+  - Pattern metadata tracking (timestamps, renewal counts)
+  - Removal history and analytics
+
+**Previous Release:** v0.9.1 (2025-10-05)
+- **AI Analysis Settings (v0.9.0)** - Performance modes and configurable AI features
+  - Three performance modes: Fast, Balanced, Comprehensive
+  - Five individually configurable AI features
+  - Complete control over speed vs intelligence tradeoff
+  - New configuration category in `adf config`
+  - Performance mode display at interview start
+- **Resume from Exit (v0.8.0)** - Resume interviews after exit
+  - Type `exit` or press Ctrl+C to save progress and quit
+  - Resume with `adf init` continues from last question
+  - Already-answered questions automatically skipped
+  - Graceful quit handling everywhere
+- **UX Improvements (v0.5.1-v0.7.1)** - Enhanced user experience
+  - Terminal input restoration (v0.7.1)
+  - Better existing project detection with clear options
+  - Post-install information display
+  - Configuration validation and auto-reset
+
+**Previous Releases:**
+- **v0.5.0** - Intelligent Answer Analysis & Dynamic Question Pipeline
+- **v0.4.36** - Multi-IDE Improvements & Config Command Enhancement
+- **v0.4.12** - Learning System (Phase 4.2) & Smart Question Filtering (Phase 4.1)
+- **v0.3.6** - Configuration Command & Optional AI Setup
+- **v0.3.4** - Multi-Provider AI Integration (Anthropic, OpenAI, Google, OpenRouter)
+- **v0.2.0** - Quality-Based Progress Tracking & Resume Capability
+- **v0.1.0** - Initial Release
 
 ## Troubleshooting
 

package/lib/frameworks/interviewer.js

@@ -278,12 +278,19 @@ class Interviewer {
       console.log('');
       projectContext = { type: 'unknown', frameworks: [], languages: [], confidence: 0 };
 
-      // Still initialize skip tracker for basic tracking
-      this.skipTracker = new SkipTracker(this.projectPath, {
-        projectType: 'unknown',
-        frameworks: [],
-        languages: []
-      });
+      // Still initialize skip tracker for basic tracking (only if pattern detection enabled)
+      if (this.analysisConfig.features.patternDetection) {
+        this.skipTracker = new SkipTracker(this.projectPath, {
+          projectType: 'unknown',
+          frameworks: [],
+          languages: []
+        });
+      }
+    }
+    } else {
+      // Smart filtering disabled - set defaults
+      console.log(chalk.gray('ℹ️  Smart filtering disabled by configuration\n'));
+      projectContext = { type: 'unknown', frameworks: [], languages: [], confidence: 0 };
     }
 
     // Get questions for framework