agentic-qe 2.1.2 → 2.2.1

package/.claude/skills/agentic-quality-engineering/SKILL.md

@@ -1,10 +1,10 @@
 ---
 name: agentic-quality-engineering
-description: "AI agents as force multipliers for quality work. Core skill for all 18 QE agents using PACT principles."
+description: "AI agents as force multipliers for quality work. Core skill for all 19 QE agents using PACT principles."
 category: qe-core
 priority: critical
 tokenEstimate: 1400
-agents: [qe-test-generator, qe-test-executor, qe-coverage-analyzer, qe-quality-gate, qe-quality-analyzer, qe-performance-tester, qe-security-scanner, qe-requirements-validator, qe-production-intelligence, qe-fleet-commander, qe-deployment-readiness, qe-regression-risk-analyzer, qe-test-data-architect, qe-api-contract-validator, qe-flaky-test-hunter, qe-visual-tester, qe-chaos-engineer, qe-code-complexity]
+agents: [qe-test-generator, qe-test-executor, qe-coverage-analyzer, qe-quality-gate, qe-quality-analyzer, qe-performance-tester, qe-security-scanner, qe-requirements-validator, qe-production-intelligence, qe-fleet-commander, qe-deployment-readiness, qe-regression-risk-analyzer, qe-test-data-architect, qe-api-contract-validator, qe-flaky-test-hunter, qe-visual-tester, qe-chaos-engineer, qe-code-complexity, qx-partner]
 implementation_status: optimized
 optimization_version: 1.0
 last_optimized: 2025-12-02
@@ -53,7 +53,7 @@ When implementing agentic QE or coordinating agents:
 | **C**ollaborative | Multi-agent coordination | Provide context |
 | **T**argeted | Risk-based prioritization | Define risk areas |
 
-### 18-Agent Fleet
+### 19-Agent Fleet
 | Category | Agents | Primary Use |
 |----------|--------|-------------|
 | Core Testing (5) | test-generator, test-executor, coverage-analyzer, quality-gate, quality-analyzer | Daily testing |
@@ -76,7 +76,7 @@ Sequential: risk-analyzer → test-gen → executor → coverage → gate
 ✅ Coverage gaps detected in real-time
 ✅ Bugs caught pre-production
 ❌ Agents acting without human oversight on critical decisions
-❌ Deploying all 18 agents at once (start with 1-2)
+❌ Deploying all 19 agents at once (start with 1-2)
 
 ---
 

package/.claude/skills/cicd-pipeline-qe-orchestrator/README.md

@@ -4,7 +4,7 @@
 
 ## What is This?
 
-The CI/CD Pipeline QE Orchestrator is an advanced Claude Code skill that provides intelligent, phase-based quality engineering across your entire software delivery pipeline. It intelligently selects from 37 QE skills and coordinates 18 specialized QE agents to ensure holistic quality coverage from commit to production.
+The CI/CD Pipeline QE Orchestrator is an advanced Claude Code skill that provides intelligent, phase-based quality engineering across your entire software delivery pipeline. It intelligently selects from 41 QE skills and coordinates 19 specialized QE agents to ensure holistic quality coverage from commit to production.
 
 ## Quick Start
 
@@ -85,7 +85,7 @@ Automatically selects optimal skills based on:
 
 ### 🤖 Agent Coordination
 
-Orchestrates 18 specialized QE agents:
+Orchestrates 19 specialized QE agents:
 - Parallel execution for speed
 - Sequential execution for dependencies
 - Memory-based coordination via `aqe/*` namespace
@@ -129,9 +129,9 @@ Adapts testing approach based on:
 
 ## Integration with AQE Fleet
 
-### All 37 Skills Available
+### All 41 Skills Available
 
-The orchestrator can invoke any of the 37 QE skills:
+The orchestrator can invoke any of the 41 QE skills:
 
 **Phase 1 Skills (18)**:
 - Core Testing: agentic-quality-engineering, context-driven-testing, holistic-testing-pact
@@ -145,12 +145,15 @@ The orchestrator can invoke any of the 37 QE skills:
 - Specialized: accessibility-testing, mobile-testing, database-testing, contract-testing, chaos-engineering-resilience, compatibility-testing, localization-testing, compliance-testing, visual-testing-advanced
 - Infrastructure: test-environment-management, test-reporting-analytics
 
-**Phase 3 Skills (3)**:
-- Strategic: six-thinking-hats, brutal-honesty-review, cicd-pipeline-qe-orchestrator
+**Phase 3 Skills (4)**:
+- Strategic: six-thinking-hats, brutal-honesty-review, sherlock-review, cicd-pipeline-qe-orchestrator
 
-### All 18 Agents Available
+**Phase 4 Skills (3)**:
+- Advanced: testability-scoring, verification-quality, pair-programming
 
-The orchestrator coordinates all 18 QE agents:
+### All 19 Agents Available
+
+The orchestrator coordinates all 19 QE agents:
 
 **Core Testing (5)**: test-generator, test-executor, coverage-analyzer, quality-gate, quality-analyzer
 **Performance & Security (2)**: performance-tester, security-scanner
@@ -277,8 +280,8 @@ See main SKILL.md [Troubleshooting section](SKILL.md#troubleshooting) for:
 - [Microservice Pipeline Workflow](resources/workflows/microservice-pipeline.md)
 - [Monolith Pipeline Workflow](resources/workflows/monolith-pipeline.md)
 - [Mobile Pipeline Workflow](resources/workflows/mobile-pipeline.md)
-- [All 37 QE Skills Reference](https://github.com/proffesor-for-testing/agentic-qe/blob/main/docs/reference/skills.md)
-- [All 18 QE Agents Reference](https://github.com/proffesor-for-testing/agentic-qe/blob/main/docs/reference/agents.md)
+- [All 41 QE Skills Reference](https://github.com/proffesor-for-testing/agentic-qe/blob/main/docs/reference/skills.md)
+- [All 19 QE Agents Reference](https://github.com/proffesor-for-testing/agentic-qe/blob/main/docs/reference/agents.md)
 
 ## Contributing
 
@@ -298,4 +301,4 @@ Part of the Agentic QE Fleet - MIT License
 
 **Created**: 2025-11-13
 **Version**: 1.0.0
-**Integrations**: 37 QE Skills, 18 QE Agents, All CI/CD platforms
+**Integrations**: 41 QE Skills, 19 QE Agents, All CI/CD platforms

package/.claude/skills/skills-manifest.json

@@ -84,7 +84,7 @@
     "agentic-quality-engineering": {
       "id": "agentic-quality-engineering",
       "name": "Agentic Quality Engineering",
-      "description": "Core foundational skill for all 18 QE agents with PACT principles (Proactive, Autonomous, Collaborative, Targeted).",
+      "description": "Core foundational skill for all 19 QE agents with PACT principles (Proactive, Autonomous, Collaborative, Targeted).",
       "category": "qe-core",
       "priority": "critical",
       "file": "agentic-quality-engineering/SKILL.md",
@@ -95,7 +95,7 @@
       "lastOptimized": "2025-12-02",
       "tags": ["agents", "pact", "core", "foundation", "autonomous-testing"],
       "dependencies": [],
-      "agents": ["qe-test-generator", "qe-test-executor", "qe-coverage-analyzer", "qe-quality-gate", "qe-quality-analyzer", "qe-performance-tester", "qe-security-scanner", "qe-requirements-validator", "qe-production-intelligence", "qe-fleet-commander", "qe-deployment-readiness", "qe-regression-risk-analyzer", "qe-test-data-architect", "qe-api-contract-validator", "qe-flaky-test-hunter", "qe-visual-tester", "qe-chaos-engineer", "qe-code-complexity"]
+      "agents": ["qe-test-generator", "qe-test-executor", "qe-coverage-analyzer", "qe-quality-gate", "qe-quality-analyzer", "qe-performance-tester", "qe-security-scanner", "qe-requirements-validator", "qe-production-intelligence", "qe-fleet-commander", "qe-deployment-readiness", "qe-regression-risk-analyzer", "qe-test-data-architect", "qe-api-contract-validator", "qe-flaky-test-hunter", "qe-visual-tester", "qe-chaos-engineer", "qe-code-complexity", "qx-partner"]
     },
     "holistic-testing-pact": {
       "id": "holistic-testing-pact",

package/CHANGELOG.md

@@ -7,6 +7,144 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [2.2.1] - 2025-12-07
+
+### Fixed
+
+#### Database Persistence Unification (Issue #118)
+- **Unified database to single `memory.db`**: Fixed database fragmentation where data was scattered across 3 files (memory.db, swarm-memory.db, agentdb.db)
+- **Fixed CLI data visibility**: `aqe learn status` and `aqe patterns list` now query actual tables (`learning_experiences`, `patterns`, `q_values`) instead of `memory_entries`
+- **Added `queryRaw()` method**: New public method on SwarmMemoryManager for direct table queries
+- **Deprecated AgentDB**: Marked for removal in v3.0.0 with proper warnings
+
+### Changed
+- All persistence now uses `getSharedMemoryManager()` / `initializeSharedMemoryManager()` singleton pattern
+- Removed default `agentdb.db` path creation from agent factory
+- CLI commands (learn, improve, patterns, routing) updated to use shared memory manager
+
+## [2.2.0] - 2025-12-06
+
+### 🧠 Self-Learning AQE Fleet Upgrade (Issue #118)
+
+Major release introducing reinforcement learning algorithms, cross-agent experience sharing, dependency injection, and LLM provider abstraction - enabling agents to learn from each other and persist knowledge across sessions.
+
+### Added
+
+#### Reinforcement Learning Algorithms (`src/learning/algorithms/`)
+- **AbstractRLLearner**: Base class for all RL algorithms with common interfaces
+- **SARSALearner**: On-policy temporal difference learning algorithm
+  - ε-greedy exploration with decay
+  - Configurable learning rate and discount factor
+  - State-action value function updates
+- **ActorCriticLearner (A2C)**: Combined policy and value learning
+  - Policy network (actor) with softmax action selection
+  - Value network (critic) for state evaluation
+  - Advantage-based policy updates with entropy regularization
+- **PPOLearner**: Proximal Policy Optimization
+  - Clipped surrogate objective for stable updates
+  - GAE (Generalized Advantage Estimation)
+  - Mini-batch training with multiple epochs
+  - Adaptive KL penalty mechanism
+- **Algorithm Switching**: Dynamic switching between Q-Learning, SARSA, A2C, and PPO via `switchAlgorithm()`
+
+#### Experience Sharing Protocol (`src/learning/ExperienceSharingProtocol.ts`)
+- **Gossip-based P2P protocol**: Agents share successful experiences with peers
+- **Priority-based sharing**: High-value experiences propagated first
+- **Conflict resolution**: Vector clocks for handling concurrent updates
+- **Transfer learning discount**: 0.5 factor for shared vs local experiences
+- **Event-driven integration**: `experience_received` events trigger cross-agent learning
+- **Sharing statistics**: Track experiences shared, received, peer connections
+
+#### LLM Provider Abstraction (`src/providers/`)
+- **ILLMProvider interface**: Common interface for all LLM providers
+  - `complete()`, `streamComplete()`, `embed()`, `countTokens()`
+  - `healthCheck()`, `getMetadata()`, `shutdown()`
+  - Cost tracking and usage statistics
+- **ClaudeProvider**: Anthropic Claude API integration
+  - Prompt caching support (reduced costs)
+  - Token counting via API
+  - Streaming completions
+- **RuvllmProvider**: Local LLM server integration
+  - Zero-cost local inference
+  - OpenAI-compatible API
+  - Embeddings support (optional)
+- **LLMProviderFactory**: Multi-provider orchestration
+  - Automatic fallback on provider failure
+  - Health monitoring with configurable intervals
+  - Best provider selection by criteria (cost, capability, location)
+  - Hybrid router for transparent multi-provider usage
+
+#### Dependency Injection System (`src/core/di/`)
+- **DIContainer**: Lightweight IoC container
+  - Singleton, factory, and instance scopes
+  - Lazy initialization support
+  - Constructor and factory injection
+- **AgentDependencies**: Agent-specific DI management
+  - `withDI()` mixin pattern for agents
+  - Automatic service resolution
+  - Lifecycle management (initialize, dispose)
+- **Service registration**: LearningEngine, MemoryCoordinator, providers
+
+#### Distributed Pattern Library (`src/memory/`)
+- **DistributedPatternLibrary**: Cross-agent pattern storage
+- **PatternQualityScorer**: ML-based pattern ranking
+- **PatternReplicationService**: Pattern synchronization across agents
+
+#### LearningEngine Integration
+- Extended config: `enableExperienceSharing`, `experienceSharingPriority`
+- New methods:
+  - `enableExperienceSharing(protocol)`: Activate cross-agent sharing
+  - `shareExperienceWithPeers(experience, priority)`: Manual sharing
+  - `handleReceivedExperience(experienceId)`: Process incoming experiences
+  - `queryPeerExperiences(query)`: Search peer knowledge
+  - `getExperienceSharingStats()`: Retrieve sharing metrics
+- Auto-sharing: Successful executions automatically shared with peers
+
+### Changed
+
+- **QLearning**: Refactored to use AbstractRLLearner base class
+- **LearningEngine**: Integrated ExperienceSharingProtocol with event listeners
+- **types.ts**: Added RLAlgorithmType, ExtendedLearningConfig, sharing-related types
+- **index.ts**: Exported all new RL algorithms and experience sharing components
+
+### Tests Added
+
+#### Provider Tests (`tests/providers/`)
+- **ClaudeProvider.test.ts**: 21 tests covering initialization, completion, streaming, cost tracking, health checks
+- **RuvllmProvider.test.ts**: 20 tests for local LLM provider including embeddings
+- **LLMProviderFactory.test.ts**: 27 tests for multi-provider orchestration
+
+#### Algorithm Tests (`tests/learning/`)
+- **SARSALearner.test.ts**: 12 tests for on-policy TD learning
+- **ActorCriticLearner.test.ts**: 15 tests for A2C algorithm
+- **PPOLearner.test.ts**: 18 tests for PPO including GAE and clipping
+- **AlgorithmSwitching.test.ts**: 8 tests for dynamic algorithm changes
+- **ExperienceSharingProtocol.test.ts**: 36 tests for P2P experience sharing
+
+#### DI Tests (`tests/core/di/`)
+- **DIContainer.test.ts**: 47 tests for IoC container functionality
+- **AgentDependencies.test.ts**: 15 tests for agent DI mixin
+
+#### Memory Tests (`tests/memory/`)
+- **DistributedPatternLibrary.test.ts**: Pattern storage tests
+- **PatternQualityScorer.test.ts**: ML scoring tests
+- **PatternReplicationService.test.ts**: Replication tests
+- **integration/**: End-to-end memory integration tests
+
+### Performance
+
+| Metric | Before | After | Notes |
+|--------|--------|-------|-------|
+| RL Algorithms | 1 (Q-Learning) | 4 (+SARSA, A2C, PPO) | 4x algorithm options |
+| Cross-Agent Learning | None | Full P2P | Agents share experiences |
+| Provider Flexibility | Claude only | Claude + Local | Cost-free local option |
+| Test Coverage | ~150 tests | ~250 tests | +100 new tests |
+
+### References
+
+- Issue: [#118 - Self-Learning AQE Fleet Upgrade](https://github.com/proffesor-for-testing/agentic-qe/issues/118)
+- Related: Learning System Phase 2 (Milestone 2.2)
+
 ## [2.1.2] - 2025-12-06
 
 ### 🚀 MCP Tools Optimization - 87% Context Reduction (Issue #115)