agentic-qe
Advanced tools
Agentic Quality Engineering V3 - Domain-Driven Design Architecture with 13 Bounded Contexts, O(log n) coverage analysis, ReasoningBank learning, 60 specialized QE agents, mathematical Coherence verification, deep Claude Flow integration
Release Notes | Changelog | Contributors | Issues | Discussions
V3 brings Domain-Driven Design architecture, 13 bounded contexts, 60 specialized QE agents, TinyDancer intelligent model routing, ReasoningBank learning with Dream cycles, HNSW vector search, mathematical Coherence verification, full MinCut/Consensus integration across all 13 domains, RVF cognitive container integration with portable brain export/import, 11 coding agent platform integrations (Claude Code, OpenCode, Kiro, GitHub Copilot, Cursor, Cline, Kilo Code, Roo Code, OpenAI Codex CLI, Windsurf, Continue.dev), and deep integration with Claude Flow and Agentic Flow.
🏗️ DDD Architecture | 🧠 ReasoningBank + Dream Cycles | 🎯 TinyDancer Model Routing | 🔍 HNSW Vector Search | 👑 Queen Coordinator | 📊 O(log n) Coverage | 🔗 Claude Flow Integration | 🎯 13 Bounded Contexts | 📚 78 QE Skills | 🧬 Coherence Verification | ✅ Trust Tiers | 🛡️ Governance | 🌐 11 Platform Integrations
# Install globally
npm install -g agentic-qe
# Initialize your project (interactive mode)
cd your-project
aqe init
# Or with auto-configuration (no prompts, configures MCP automatically)
aqe init --auto
# Include OpenCode assets (agents, skills, tools, permissions)
aqe init --auto --with-opencode
# Include AWS Kiro IDE assets (agents, skills, hooks, steering)
aqe init --auto --with-kiro
# Include any coding agent platform (Copilot, Cursor, Cline, etc.)
aqe init --auto --with-copilot --with-cursor --with-cline
# Or include all 8 additional platforms at once
aqe init --auto --with-all-platforms
Note:
aqe initautomatically configures the MCP server in.mcp.json— Claude Code will auto-start it when connecting. For standalone MCP server usage (non-Claude-Code clients), runaqe-mcpornpx agentic-qe mcp.
AQE is exposed as an MCP server and can be used from any client that supports MCP tool connections.
# For Claude Code: aqe init --auto configures .mcp.json automatically
# Claude Code will auto-start the MCP server on connection
# For OpenCode: provision assets automatically during init
aqe init --auto --with-opencode # installs agents, skills, tools, permissions, opencode.json
# For AWS Kiro: provision Kiro-native assets during init
aqe init --auto --with-kiro # installs .kiro/ agents, skills, hooks, steering, MCP config
# For GitHub Copilot, Cursor, Cline, Kilo Code, Roo Code, Codex, Windsurf, Continue.dev:
aqe init --auto --with-copilot # .vscode/mcp.json + copilot-instructions.md
aqe init --auto --with-cursor # .cursor/mcp.json + .cursorrules
aqe init --auto --with-cline # cline_mcp_settings.json + custom QE mode
aqe init --auto --with-all-platforms # all 8 platforms at once
# Or set up a specific platform after init:
aqe platform setup copilot # configure a single platform
aqe platform list # show all platforms with install status
aqe platform verify cursor # validate config format and content
# For other MCP clients: start the server manually
aqe-mcp # if installed globally
npx agentic-qe mcp # without global install
For client-specific setup examples, see Platform Setup Guide.
What V3 provides:
aqe/* namespacesaqe init --with-opencode auto-provisioningaqe init --with-kiro auto-provisioningaqe platform list/setup/verify CLI# 1. Install
npm install -g agentic-qe
# 2. Initialize (auto-detects your project, enables all 13 domains, configures MCP)
cd your-project && aqe init --auto
# 3. Use from Claude Code — MCP tools are available immediately (prefix: mcp__agentic-qe__)
# Or start MCP server manually for other clients: aqe-mcp
What happens:
.mcp.json — Claude Code auto-connects on next session| Problem | AQE Solution |
|---|---|
| Writing comprehensive tests is tedious and time-consuming | AI agents generate tests automatically with pattern reuse across projects |
| Test suites become slow and expensive at scale | Sublinear O(log n) algorithms for coverage analysis and intelligent test selection |
| Flaky tests waste developer time debugging false failures | ML-powered detection with root cause analysis and fix recommendations |
| AI testing tools are expensive | TinyDancer 3-tier model routing reduces costs by matching task complexity to appropriate model |
| No memory between test runs—every analysis starts from scratch | ReasoningBank remembers patterns, strategies, and what works for your codebase |
| Agents waste tokens reading irrelevant code | Code Intelligence provides token reduction with semantic search and knowledge graphs |
| Quality engineering requires complex coordination | Queen Coordinator orchestrates 60 agents across 13 domains with consensus and MinCut topology |
| Tools don't understand your testing frameworks | Works with Jest, Cypress, Playwright, Vitest, Mocha, Jasmine, AVA |
V3 is built on 13 DDD Bounded Contexts, each with dedicated agents and clear responsibilities:
| Domain | Purpose | Key Agents |
|---|---|---|
| test-generation | AI-powered test creation | qe-test-architect, qe-tdd-specialist |
| test-execution | Parallel execution & retry | qe-parallel-executor, qe-retry-handler |
| coverage-analysis | O(log n) gap detection | qe-coverage-specialist, qe-gap-detector |
| quality-assessment | Quality gates & decisions | qe-quality-gate, qe-risk-assessor |
| defect-intelligence | Prediction & root cause | qe-defect-predictor, qe-root-cause-analyzer |
| requirements-validation | BDD & testability | qe-requirements-validator, qe-bdd-generator |
| code-intelligence | Knowledge graph & search | qe-code-intelligence, qe-kg-builder |
| security-compliance | SAST/DAST & audit | qe-security-scanner, qe-security-auditor |
| contract-testing | API contracts & GraphQL | qe-contract-validator, qe-graphql-tester |
| visual-accessibility | Visual regression & a11y | qe-visual-tester, qe-accessibility-auditor |
| chaos-resilience | Chaos engineering & load | qe-chaos-engineer, qe-load-tester |
| learning-optimization | Cross-domain learning | qe-learning-coordinator, qe-pattern-learner |
| enterprise-integration | SOAP, SAP, ESB, OData | qe-soap-tester, qe-sap-rfc-tester, qe-sod-analyzer |
AQE includes 78 QE skills (46 Tier 3 verified + 32 additional). Trust tiers apply to core QE skills:
| Tier | Badge | Count | Description |
|---|---|---|---|
| Tier 3 - Verified |  | 46 | Full evaluation test suite |
| Tier 2 - Validated |  | 7 | Has executable validator |
| Tier 1 - Structured |  | 5 | Has JSON output schema |
| Tier 0 - Advisory |  | 5 | SKILL.md guidance only |
Tier 3 Skills are recommended for production use - they have:
# Check skill trust tier
aqe eval status --skill security-testing
# Run skill evaluation
aqe eval run --skill security-testing --model claude-sonnet-4
# View all trust tiers
cat .claude/skills/TRUST-TIERS.md
[Full documentation: docs/guides/skill-validation.md]
V3.1.0 adds full browser automation support via @claude-flow/browser integration:
| Component | Description |
|---|---|
| BrowserSwarmCoordinator | Parallel multi-viewport testing (4x faster) |
| BrowserSecurityScanner | URL validation, PII detection with auto-masking |
| 9 Workflow Templates | YAML-based reusable browser workflows |
| TrajectoryAdapter | SONA learning integration with HNSW indexing |
Available Workflow Templates:
login-flow, oauth-flow - Authentication testingform-validation, navigation-flow - User journey testingvisual-regression, accessibility-audit - Quality validationperformance-audit, api-integration, scraping-workflow - Advanced workflows# Use browser automation from Claude Code
claude "Use security-visual-testing skill to test https://example.com across mobile, tablet, desktop viewports"
# Load and execute a workflow template
aqe workflow run login-flow --vars '{"username": "test", "password": "secret"}'
The qe-queen-coordinator manages the entire fleet with intelligent task distribution:
qe-queen-coordinator
(Queen)
|
+--------------------+--------------------+
| | |
TEST DOMAIN QUALITY DOMAIN LEARNING DOMAIN
(test-generation) (quality-assessment) (learning-optimization)
| | |
- test-architect - quality-gate - learning-coordinator
- tdd-specialist - risk-assessor - pattern-learner
- integration-tester - deployment-advisor - transfer-specialist
Capabilities:
claude "Use qe-queen-coordinator to orchestrate release validation for v2.1.0 with 90% coverage target"
The Queen Coordinator is extended with Agent Teams (ADR-064) for hybrid fleet communication:
| Feature | Description |
|---|---|
| Mailbox Messaging | Direct agent-to-agent and domain-scoped broadcast messaging |
| Distributed Tracing | TraceContext propagation across messages for end-to-end task visibility |
| Dynamic Scaling | Workload-based auto-scaling with configurable policies and cooldowns |
| Competing Hypotheses | Multi-agent root cause investigation with evidence scoring, auto-triggered on critical failures |
| Federation | Cross-service routing with health monitoring and service discovery |
| Circuit Breakers | Per-domain fault isolation with automatic recovery |
| Task DAG | Topological ordering with cycle detection for multi-step workflows |
Fleet Tiers — Activate the level of coordination your project needs:
| Tier | Agents | Best For |
|---|---|---|
| Lite | 1-4 | Small projects, focused tasks |
| Standard | 5-10 | Team projects, multi-domain coordination |
| Full | 11-15 | Enterprise, cross-fleet federation |
claude "Use qe-queen-coordinator with agent teams to investigate flaky test failures across test-execution and defect-intelligence domains"
V3 agents learn and improve through the ReasoningBank pattern storage:
| Component | Description |
|---|---|
| Experience Storage | Store successful patterns with confidence scores |
| HNSW Indexing | Fast O(log n) similarity search for pattern matching |
| Experience Replay | Learn from past successes and failures |
| Cross-Project Transfer | Share patterns between projects |
# Search learned patterns
aqe hooks search --query "test patterns"
# View learning statistics
aqe learning stats
V3.7.7 adds 7 adversarial quality features inspired by loki-mode — designed to catch sycophantic AI outputs, hollow tests, and routing drift. All features are enabled by default (opt-out via config flags).
| Feature | Config Flag | Description |
|---|---|---|
| Anti-Sycophancy Scorer | enableSycophancyCheck | Detects rubber-stamp consensus via Jaccard similarity, confidence uniformity, and reasoning overlap |
| Test Quality Gates | enableTestQualityGate | Catches tautological assertions (expect(true).toBe(true)), empty test bodies, and missing source imports |
| Blind Review | N/A (API option) | Runs N parallel test generators with varied temperatures, deduplicates via Jaccard |
| EMA Calibration | enableEMACalibration | Exponential moving average tracks per-agent accuracy, derives dynamic voting weights |
| Edge-Case Injection | enableEdgeCaseInjection | Queries historical patterns and injects proven edge cases into test generation prompts |
| Complexity Composition | enableComplexityComposition | Maps 8-dimension complexity (AST + security + concurrency) to agent team composition |
| Auto-Escalation | enableAutoEscalation | Consecutive failures auto-promote agent tier; consecutive successes auto-demote |
// All features are ON by default. To disable specific features:
const config: Partial<RoutingConfig> = {
enableEMACalibration: false, // Disable EMA voting weights
enableAutoEscalation: false, // Disable auto tier promotion
};
const consensusConfig: Partial<ConsensusEngineConfig> = {
enableSycophancyCheck: false, // Disable rubber-stamp detection
};
const testConfig: Partial<TestGeneratorConfig> = {
enableTestQualityGate: false, // Disable tautology detection
enableEdgeCaseInjection: false, // Disable pattern injection
};
See docs/loki-mode-features.md for detailed usage examples and configuration reference.
V3 introduces Dream cycles for neural consolidation and continuous improvement:
| Feature | Description |
|---|---|
| Dream Cycles | Background neural consolidation (30s max) with spreading activation |
| 9 RL Algorithms | Q-Learning, SARSA, DQN, PPO, A2C, DDPG, Actor-Critic, Policy Gradient, Decision Transformer |
| SONA Integration | Self-Optimizing Neural Architecture with <0.05ms adaptation |
| Novelty Scoring | Prioritize learning from novel patterns |
| Concept Graphs | Build semantic connections between quality patterns |
# Trigger dream cycle for pattern consolidation
aqe learning dream
# View learning dashboard and trajectory
aqe learning dashboard
V3.7 integrates RVF (RuVector Format) cognitive container capabilities across 4 workstreams:
| Feature | Description |
|---|---|
| MinCut Task Routing | Models task complexity as a graph problem using vertex connectivity (lambda) for intelligent 3-tier routing |
| RVCOW Dream Branching | Copy-on-write branches for safe dream cycle experimentation — speculative insights are isolated until merged |
| Cryptographic Witness Chain | SHA-256 hash-chained audit trail for quality gate decisions, pattern promotions, and test completions |
| Unified HNSW Search | Consolidated 3 fragmented implementations behind a single progressive adapter (flat scan for small, full HNSW for large collections) |
| Brain Export/Import | Portable QE intelligence containers — export patterns, Q-values, and insights for sharing across environments |
| MinCut Test Optimizer | Identifies critical vs skippable tests using graph-based coverage analysis for faster test suites |
| RVF Dual-Writer | Best-effort dual-write to SQLite + RVF containers, preparing for future native RVF storage |
# Export your QE brain as a portable container
aqe brain export --output ./my-brain
# Import a brain into another environment
aqe brain import --input ./my-brain --dry-run
# View brain statistics
aqe brain info --input ./my-brain
V3.7.2 adds full AWS Kiro IDE integration, converting AQE agents, skills, hooks, and steering files to Kiro-native formats via aqe init --with-kiro.
| Asset | Count | Description |
|---|---|---|
| Agents | 87 | .kiro/agents/*.json — 62 QE + 15 n8n + 10 infra/testing agents |
| Skills | 86 | .kiro/skills/*/SKILL.md — full instructional content from Claude Code sources |
| Hooks | 5 | .kiro/hooks/*.kiro.hook — event-driven automation (test update, coverage, security, quality gate, pre-commit) |
| Steering | 2 | .kiro/steering/*.md — QE standards and testing conventions |
| MCP Config | 1 | .kiro/settings/mcp.json — auto-approved AQE tools |
Setup:
# Provision Kiro assets during init
aqe init --auto --with-kiro
# Or add Kiro to an existing project
aqe init --with-kiro
What gets generated:
.kiro/
├── agents/ # 87 agent definitions (Kiro JSON format)
├── skills/*/SKILL.md # 86 skill definitions (full markdown content)
├── hooks/ # 5 event-driven hooks
│ ├── aqe-test-updater.kiro.hook # Auto-update tests on source changes
│ ├── aqe-coverage-check.kiro.hook # Coverage analysis on test edits
│ ├── aqe-security-scan.kiro.hook # Security scan on auth file changes
│ ├── aqe-spec-quality-gate.kiro.hook # Quality gate after spec tasks
│ └── aqe-pre-commit-quality.kiro.hook # Quality check before agent stops
├── steering/ # 2 project-wide guidance files
│ ├── qe-standards.md # QE standards (auto-triggered)
│ └── testing-conventions.md # Test file conventions (fileMatch)
└── settings/mcp.json # MCP server config with auto-approved tools
Key features:
mcp:agentic-qe: references to Kiro's @agentic-qe/ formatbash→shell, edit→write).kiro/ directory in --auto mode--upgrade is usedV3.7.4 adds support for 8 additional coding agent platforms, bringing the total to 11 supported platforms. A single AQE MCP server works across all of them.
| Platform | Users | Config Format | Setup Flag |
|---|---|---|---|
| Claude Code | Native | .mcp.json | Built-in |
| OpenCode | Growing | opencode.json | --with-opencode |
| AWS Kiro | Growing | .kiro/settings/mcp.json | --with-kiro |
| GitHub Copilot | 20M | .vscode/mcp.json | --with-copilot |
| Cursor | 360K paid | .cursor/mcp.json | --with-cursor |
| Cline | 5M installs | cline_mcp_settings.json | --with-cline |
| Kilo Code | 1.5M | .kilocode/mcp.json | --with-kilocode |
| Roo Code | Growing | .roo/mcp.json | --with-roocode |
| OpenAI Codex CLI | 1M+ | .codex/config.toml | --with-codex |
| Windsurf | Large | .windsurf/mcp_config.json | --with-windsurf |
| Continue.dev | Enterprise | .continue/config.yaml | --with-continuedev |
Each installer generates:
.cursorrules, copilot-instructions.md, AGENTS.md)# Set up all platforms at once
aqe init --auto --with-all-platforms
# Or individual platforms
aqe platform setup copilot
aqe platform setup cursor
# Check what's installed
aqe platform list
# Validate a platform's config
aqe platform verify cursor
For detailed per-platform setup instructions, see Platform Setup Guide.
TinyDancer (ADR-026) provides 3-tier intelligent model routing for cost optimization:
| Complexity Score | Model | Use Cases |
|---|---|---|
| 0-20 (Simple) | Haiku | Syntax fixes, type additions, simple refactors |
| 20-70 (Moderate) | Sonnet | Bug fixes, test generation, code review |
| 70+ (Critical) | Opus | Architecture, security, complex reasoning |
Routing Features:
# Route a task through TinyDancer
aqe llm route --task "fix type errors in user-service.ts"
# View routing cost analysis
aqe llm cost
V3.3.5 unifies cross-phase feedback loops with UnifiedMemoryManager:
.agentic-qe/memory.dbqcsd/strategic, qcsd/tactical, qcsd/operational, qcsd/quality-criteriaV3.4.0 adds support for industry-standard agent communication protocols:
| Protocol | Standard | Purpose |
|---|---|---|
| AG-UI | Anthropic | Agent-to-User streaming interface with lifecycle events |
| A2A | Agent-to-Agent interoperability with task/artifact exchange | |
| A2UI | Hybrid | Unified UI components combining streaming + events |
Programmatic usage:
import { AGUIAdapter, A2AAdapter } from 'agentic-qe';
// AG-UI: Stream test generation progress to UI
const agui = new AGUIAdapter();
await agui.streamTask({
type: 'test-generation',
onProgress: (event) => updateProgressBar(event.progress),
onArtifact: (test) => displayGeneratedTest(test),
});
// A2A: Inter-agent task delegation
const a2a = new A2AAdapter();
await a2a.sendTask({
from: 'qe-test-architect',
to: 'qe-security-scanner',
task: { type: 'review-tests', files: generatedTests },
});
Benefits:
V3.3.3 achieves full MinCut/Consensus integration across all 13 domains:
| Feature | Description |
|---|---|
| Byzantine Consensus | Fault-tolerant voting for critical quality decisions |
| MinCut Topology | Graph-based self-healing agent coordination |
| Multi-Model Voting | Aggregate decisions from multiple model tiers |
| Claim Verification | Cryptographic verification of agent work claims |
| 13/13 Domain Integration | All domains use verifyFinding() for consensus |
| Topology-Aware Routing | Routes tasks avoiding weak network vertices |
| Self-Healing Triggers | shouldPauseOperations() for automatic recovery |
# Check fleet health (includes consensus and topology status)
aqe fleet status
# Via MCP tools (from Claude Code)
# mcp__agentic-qe__fleet_health({ includeTopology: true })
V3.3.0 introduces mathematical coherence verification using Prime Radiant WASM engines:
| Feature | Description |
|---|---|
| Contradiction Detection | Sheaf cohomology identifies conflicting requirements before test generation |
| Collapse Prediction | Spectral analysis predicts swarm failures before they happen |
| Causal Verification | Distinguishes true causation from spurious correlations |
| Auto-Tuning Thresholds | EMA-based calibration adapts to your codebase |
Compute Lanes - Automatic routing based on coherence energy:
| Coherence Energy | Action | Latency |
|---|---|---|
| < 0.1 (Reflex) | Execute immediately | <1ms |
| 0.1-0.4 (Retrieval) | Fetch more context | ~10ms |
| 0.4-0.7 (Heavy) | Deep analysis | ~100ms |
| > 0.7 (Human) | Escalate to Queen | Async |
Benefits:
# Coherence verification is available via MCP tools and programmatic API:
# mcp__agentic-qe__quality_assess({ scope: "coherence", includeMetrics: true })
# Verify learned patterns for consistency
aqe learning verify
Efficient coverage gap detection using Johnson-Lindenstrauss algorithms:
claude "Use qe-coverage-specialist to analyze gaps in src/ with risk scoring"
V3 deeply integrates with Claude Flow for:
aqe/* namespaces# Initialize swarm with Claude Flow
npx @claude-flow/cli@latest swarm init --topology hierarchical-mesh
# Spawn V3 agents
npx @claude-flow/cli@latest agent spawn -t qe-test-architect --name test-gen
| Category | Count | Highlights |
|---|---|---|
| Main QE Agents | 53 | Test generation, coverage, security, performance, accessibility, enterprise integration, pentest validation |
| TDD Subagents | 7 | RED/GREEN/REFACTOR with code review |
V2 Backward Compatibility: All V2 agents map to V3 equivalents automatically.
| Agent | Domain | Purpose |
|---|---|---|
| qe-queen-coordinator | coordination | Hierarchical fleet orchestration |
| qe-test-architect | test-generation | AI-powered test creation |
| qe-tdd-specialist | test-generation | TDD workflow coordination |
| qe-parallel-executor | test-execution | Multi-worker test execution |
| qe-retry-handler | test-execution | Intelligent retry with backoff |
| qe-coverage-specialist | coverage-analysis | O(log n) coverage analysis |
| qe-gap-detector | coverage-analysis | Risk-weighted gap detection |
| qe-quality-gate | quality-assessment | Quality threshold validation |
| qe-risk-assessor | quality-assessment | Multi-factor risk scoring |
| qe-deployment-advisor | quality-assessment | Go/no-go deployment decisions |
| qe-defect-predictor | defect-intelligence | ML-powered defect prediction |
| qe-root-cause-analyzer | defect-intelligence | Systematic root cause analysis |
| qe-flaky-hunter | defect-intelligence | Flaky test detection & fix |
| qe-requirements-validator | requirements-validation | Testability analysis |
| qe-bdd-generator | requirements-validation | Gherkin scenario generation |
| qe-code-intelligence | code-intelligence | Semantic code search |
| qe-kg-builder | code-intelligence | Knowledge graph construction |
| qe-dependency-mapper | code-intelligence | Dependency analysis |
| qe-security-scanner | security-compliance | SAST/DAST scanning |
| qe-security-auditor | security-compliance | Security audit & compliance |
| qe-contract-validator | contract-testing | API contract validation |
| qe-graphql-tester | contract-testing | GraphQL testing |
| qe-visual-tester | visual-accessibility | Visual regression testing |
| qe-accessibility-auditor | visual-accessibility | WCAG compliance testing |
| qe-responsive-tester | visual-accessibility | Cross-viewport testing |
| qe-chaos-engineer | chaos-resilience | Controlled fault injection |
| qe-load-tester | chaos-resilience | Load & performance testing |
| qe-performance-tester | chaos-resilience | Performance validation |
| qe-learning-coordinator | learning-optimization | Fleet-wide learning |
| qe-pattern-learner | learning-optimization | Pattern discovery |
| qe-transfer-specialist | learning-optimization | Cross-project transfer |
| qe-metrics-optimizer | learning-optimization | Hyperparameter tuning |
| qe-integration-tester | test-execution | Component integration |
| qe-mutation-tester | test-generation | Test effectiveness validation |
| qe-property-tester | test-generation | Property-based testing |
| qe-regression-analyzer | defect-intelligence | Regression risk analysis |
| qe-impact-analyzer | code-intelligence | Change impact assessment |
| qe-code-complexity | code-intelligence | Complexity metrics |
| qe-qx-partner | quality-assessment | QA + UX collaboration |
| qe-fleet-commander | coordination | Large-scale orchestration |
| qe-integration-architect | code-intelligence | V3 integration design |
| qe-product-factors-assessor | quality-assessment | SFDIPOT product factors analysis |
| qe-test-idea-rewriter | test-generation | Transform passive tests to active actions |
| qe-quality-criteria-recommender | quality-assessment | HTSM v6.3 Quality Criteria analysis |
| qe-devils-advocate | quality-assessment | Adversarial review of agent outputs |
| Subagent | Phase | Purpose |
|---|---|---|
| qe-tdd-red | RED | Write failing tests |
| qe-tdd-green | GREEN | Implement minimal code |
| qe-tdd-refactor | REFACTOR | Improve code quality |
| qe-code-reviewer | REVIEW | Code quality validation |
| qe-integration-reviewer | REVIEW | Integration review |
| qe-performance-reviewer | REVIEW | Performance review |
| qe-security-reviewer | REVIEW | Security review |
claude "Use qe-queen-coordinator to run full quality assessment:
1. Generate tests for src/services/*.ts
2. Execute tests with parallel workers
3. Analyze coverage gaps with risk scoring
4. Run security scan
5. Validate quality gate at 90% threshold
6. Provide deployment recommendation"
What happens:
claude "Use qe-test-architect to create tests for PaymentService with:
- Property-based testing for validation
- 95% coverage target
- Apply learned patterns from similar services"
Output includes:
Generated 48 tests across 4 files
- unit/PaymentService.test.ts (32 unit tests)
- property/PaymentValidation.property.test.ts (8 property tests)
- integration/PaymentFlow.integration.test.ts (8 integration tests)
Coverage: 96.2%
Pattern reuse: 78% from learned patterns
Learning stored: "payment-validation-patterns" (confidence: 0.94)
claude "Use qe-tdd-specialist to implement UserAuthentication with full RED-GREEN-REFACTOR cycle"
Workflow:
claude "Coordinate security audit across the monorepo:
- qe-security-scanner for SAST/DAST
- qe-dependency-mapper for vulnerability scanning
- qe-contract-validator for API security
- qe-chaos-engineer for resilience testing"
V3 agents automatically apply relevant skills from the comprehensive QE skill library.
Core Testing & Methodologies (12)
Specialized Testing (13)
V3 Domain Skills (14)
Strategic & Communication (8)
Testing Techniques & Management (9)
n8n Workflow Testing (5) (contributed by @fndlalit)
QCSD Swarms (4) - Quality Conscious Software Delivery lifecycle
Accessibility (2)
V3 provides automatic backward compatibility with V2:
# Check migration status
aqe migrate status
# Run migration with backup
aqe migrate run --backup
# Verify migration
aqe migrate verify
What gets migrated:
| V2 Agent | V3 Agent |
|---|---|
| qe-test-generator | qe-test-architect |
| qe-coverage-analyzer | qe-coverage-specialist |
| qe-quality-gate | qe-quality-gate |
| qe-security-scanner | qe-security-scanner |
| qe-coordinator | qe-queen-coordinator |
AQE V3 supports multiple LLM providers for maximum flexibility:
| Provider | Type | Cost | Best For |
|---|---|---|---|
| Ollama | Local | FREE | Privacy, offline |
| OpenRouter | Cloud | Varies | 300+ models |
| Groq | Cloud | FREE | High-speed |
| Claude API | Cloud | Paid | Highest quality |
| Google AI | Cloud | FREE | Gemini models |
# Configure provider
export GROQ_API_KEY="gsk_..."
aqe init --auto
agentic-qe/
├── src/
│ ├── kernel/ # Shared kernel
│ ├── domains/ # 13 bounded contexts
│ │ ├── test-generation/
│ │ ├── coverage-analysis/
│ │ ├── quality-assessment/
│ │ └── ...
│ ├── routing/ # Agent routing & registry
│ ├── mcp/ # MCP server
│ └── cli/ # CLI
├── tests/ # 17,000+ tests
├── assets/agents/ # 60 QE agent definitions (53 main + 7 subagents)
├── .claude/
│ ├── agents/v3/ # Agent definitions (source)
│ └── skills/ # QE-specific skills
├── .kiro/ # AWS Kiro IDE integration
│ ├── agents/ # 87 Kiro agent definitions (JSON)
│ ├── skills/ # 86 Kiro skill definitions (SKILL.md)
│ ├── hooks/ # 5 event-driven hooks
│ ├── steering/ # 2 QE steering files
│ └── settings/ # MCP server configuration
├── docs/ # Documentation
│ ├── plans/ # Migration plans
│ ├── policies/ # Project policies
│ └── implementation/ # ADRs and DDD docs
├── package.json
└── README.md # This file
# Clone repository
git clone https://github.com/proffesor-for-testing/agentic-qe.git
cd agentic-qe
# Install V3 dependencies
cd v3
npm install
# Build
npm run build
# Run tests
npm test -- --run
| Script | Description |
|---|---|
npm run build | Compile TypeScript |
npm test -- --run | Run all tests |
npm run cli | Run CLI in dev mode |
npm run mcp | Start MCP server |
We welcome contributions! Please see CONTRIBUTING.md for details.
This project is licensed under the MIT License - see the LICENSE file for details.
Thanks to all the amazing people who have contributed to Agentic QE Fleet!
 @proffesor-for-testing Project Lead |  @fndlalit QX Partner, Testability |  @shaal Core Development |  @mondweep Architecture |
|---|
View all contributors | Become a contributor
If you find Agentic QE Fleet valuable, consider supporting its development:
| Monthly | Annual (Save $10) | |
|---|---|---|
| Price | $5/month | $50/year |
| Benefits | Sponsor recognition, Priority support | All monthly + Featured in README, Roadmap input |
| Subscribe | Monthly | Annual |
V3 is built on the shoulders of giants:
Made with ❤️ by the Agentic QE Team
FAQs
Agentic Quality Engineering V3 - Domain-Driven Design Architecture with 13 Bounded Contexts, O(log n) coverage analysis, ReasoningBank learning, 60 specialized QE agents, mathematical Coherence verification, deep Claude Flow integration
The npm package agentic-qe receives a total of 795 weekly downloads. As such, agentic-qe popularity was classified as not popular.
We found that agentic-qe demonstrated a healthy version release cadence and project activity because the last version was released less than a year ago. It has 1 open source maintainer collaborating on the project.
Did you know?
Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.
Security News
TC39’s March 2026 meeting advanced eight ECMAScript proposals, including Temporal reaching Stage 4 and securing its place in the ECMAScript 2026 specification.
Research
/Security News
Since January 31, 2026, we identified at least 72 additional malicious Open VSX extensions, including transitive GlassWorm loader extensions targeting developers.
Research
Six malicious Packagist packages posing as OphimCMS themes contain trojanized jQuery that exfiltrates URLs, injects ads, and loads FUNNULL-linked redirects.