devsquad 3.6.0__tar.gz → 3.6.5__tar.gz

{devsquad-3.6.0/devsquad.egg-info → devsquad-3.6.5}/PKG-INFO

@@ -1,24 +1,27 @@
 Metadata-Version: 2.4
 Name: devsquad
-Version: 3.6.0
-Summary: Multi-Role AI Task Orchestrator — one task in, multi-role AI collaboration, one conclusion out (Production Ready)
-Author: DevSquad Team
+Version: 3.6.5
+Summary: Production-Ready Multi-Role AI Task Orchestrator
+Author-email: DevSquad Team <devsquad@example.com>
 License: MIT
 Project-URL: Homepage, https://github.com/lulin70/DevSquad
-Project-URL: Documentation, https://github.com/lulin70/DevSquad#readme
-Project-URL: Repository, https://github.com/lulin70/DevSquad
+Project-URL: Documentation, https://github.com/lulin70/DevSquad/blob/main/README.md
+Project-URL: Repository, https://github.com/lulin70/DevSquad.git
 Project-URL: Issues, https://github.com/lulin70/DevSquad/issues
-Keywords: ai,multi-agent,orchestration,llm,collaboration
-Classifier: Development Status :: 4 - Beta
+Project-URL: PyPI, https://pypi.org/project/devsquad/
+Keywords: multi-agent,orchestration,llm,ai,collaboration,role-based,consensus,cybernetics,task-dispatch,openai,anthropic,cli,api,dashboard
+Classifier: Development Status :: 5 - Production/Stable
 Classifier: Intended Audience :: Developers
 Classifier: License :: OSI Approved :: MIT License
+Classifier: Operating System :: OS Independent
 Classifier: Programming Language :: Python :: 3
-Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
 Classifier: Topic :: Software Development :: Libraries :: Python Modules
-Requires-Python: >=3.9
+Classifier: Typing :: Typed
+Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 License-File: LICENSE
 Requires-Dist: pyyaml>=6.0
@@ -35,6 +38,7 @@ Provides-Extra: mcp
 Requires-Dist: mcp>=0.9; extra == "mcp"
 Provides-Extra: monitoring
 Requires-Dist: psutil>=5.9; extra == "monitoring"
+Requires-Dist: prometheus-client>=0.19; extra == "monitoring"
 Provides-Extra: api
 Requires-Dist: fastapi>=0.100.0; extra == "api"
 Requires-Dist: uvicorn[standard]>=0.23.0; extra == "api"
@@ -43,30 +47,68 @@ Provides-Extra: visualization
 Requires-Dist: streamlit>=1.28.0; extra == "visualization"
 Requires-Dist: jupyter>=1.0; extra == "visualization"
 Provides-Extra: alerts
-Requires-Dist: slack-sdk>=3.20; extra == "alerts"
+Requires-Dist: slack-sdk>=3.21.0; extra == "alerts"
 Provides-Extra: dev
 Requires-Dist: pytest>=7.0; extra == "dev"
 Requires-Dist: pytest-asyncio>=0.21; extra == "dev"
+Requires-Dist: pytest-cov>=4.1; extra == "dev"
+Requires-Dist: ruff>=0.4.0; extra == "dev"
+Requires-Dist: mypy>=1.0; extra == "dev"
 Requires-Dist: black>=23.0; extra == "dev"
 Requires-Dist: flake8>=6.0; extra == "dev"
-Requires-Dist: mypy>=1.0; extra == "dev"
 Provides-Extra: all
-Requires-Dist: devsquad[alerts,anthropic,api,carrymem,dev,mcp,monitoring,openai,visualization]; extra == "all"
+Requires-Dist: openai>=1.0; extra == "all"
+Requires-Dist: anthropic>=0.18; extra == "all"
+Requires-Dist: carrymem[devsquad]>=0.2.8; extra == "all"
+Requires-Dist: mcp>=0.9; extra == "all"
+Requires-Dist: psutil>=5.9; extra == "all"
+Requires-Dist: prometheus-client>=0.19; extra == "all"
+Requires-Dist: fastapi>=0.100.0; extra == "all"
+Requires-Dist: uvicorn[standard]>=0.23.0; extra == "all"
+Requires-Dist: pydantic>=2.0; extra == "all"
+Requires-Dist: streamlit>=1.28.0; extra == "all"
+Requires-Dist: jupyter>=1.0; extra == "all"
+Requires-Dist: slack-sdk>=3.21.0; extra == "all"
+Requires-Dist: pytest>=7.0; extra == "all"
+Requires-Dist: pytest-asyncio>=0.21; extra == "all"
+Requires-Dist: pytest-cov>=4.1; extra == "all"
+Requires-Dist: ruff>=0.4.0; extra == "all"
+Requires-Dist: mypy>=1.0; extra == "all"
+Requires-Dist: black>=23.0; extra == "all"
+Requires-Dist: flake8>=6.0; extra == "all"
 Dynamic: license-file
 
 # DevSquad — Multi-Role AI Task Orchestrator
 
+<details>
+<summary>📑 Table of Contents</summary>
+
+- [Features & Architecture](#-v361-cybernetics-enhancement-release)
+- [Quick Start](#-quick-start-4-ways-to-use-devsquad)
+- [Installation](#-installation)
+- [Key Features](#-key-features-v361)
+- [Cybernetics Modules](#-v361-cybernetics-module-details)
+- [Integration Architecture](#-integration-architecture)
+- [Role System](#-7-core-roles)
+- [Module Reference](#-architecture-overview-60-core-modules)
+- [CLI Usage](#-running-tests)
+- [Python API](#-configuration)
+- [Running Tests](#-running-tests)
+- [Version History](#-version-history)
+
+</details>
+
 <p align="center">
   <strong>One task → Multi-role AI collaboration → One conclusion</strong>
   <br>
-  <em>Production Ready | V3.6.0</em>
+  <em>Enterprise Ready | V3.6.5</em>
 </p>
 
 <p align="center">
   <img alt="Python" src="https://img.shields.io/badge/Python-3.9+-blue?logo=python&logoColor=white" />
   <img alt="License" src="https://img.shields.io/badge/License-MIT-green" />
-  <img alt="Tests" src="https://img.shields.io/badge/Tests-1548%20passing-brightgreen" />
-  <img alt="Version" src="https://img.shields.io/badge/V3.6.0-success" />
+  <img alt="Tests" src="https://img.shields.io/badge/Tests-1731%20passing-brightgreen" />
+  <img alt="Version" src="https://img.shields.io/badge/V3.6.5-success" />
   <img alt="CI" src="https://img.shields.io/badge/CI-GitHub_Actions-blue?logo=githubactions" />
   <img alt="Quality" src="https://img.shields.io/badge/Code%20Quality-4.3%2F5%20%E2%98%85%E2%98%85%E2%98%85%E2%98%85%E2%98%86-blue" />
   <img alt="Security" src="https://img.shields.io/badge/Security-5%2F5%20%E2%98%85%E2%98%85%E2%98%85%E2%98%85%E2%98%85-success" />
@@ -83,9 +125,196 @@ Dynamic: license-file
 
 ---
 
-## 🚀 V3.6.0: Anchor Check & Retrospective Enhancement Release
+## 🚀 V3.6.5: Enterprise Edition + E2E Testing Release
+
+**DevSquad V3.6.5** adds enterprise-grade features and comprehensive E2E testing: RBAC Engine with 15+ permissions and 5 roles, Audit Logger with SHA256 integrity chain, Multi-Tenancy Manager with 3 isolation levels, Sensitive Data Masker for PII protection, AsyncIO transformation for 2x throughput improvement, Redis Cache Integration with L1→L2→L3 three-level cache achieving 95%+ hit rate, Prometheus Monitoring with 12 core metrics, E2E Test Suite with 27 test cases covering 5 scenarios (CLI/API/Collaboration/Enterprise/Error Recovery) with 100% pass rate in 9 seconds, Code Quality improvements including print() → logging migration, TODO/FIXME cleanup, pre-commit hooks, and .editorconfig — making DevSquad truly enterprise-ready with 13,000+ lines new code, 1731 tests passing, and 97% maturity score.
+
+### 🔄 V3.6.5 Enterprise Features Details
+
+#### 1️⃣ FeedbackControlLoop (Feedback Controller)
+**Chinese Name**: 反馈闭环控制器 (Feedback Closed-Loop Controller)
+**Core Capabilities**:
+- Closed-loop feedback control with automatic iteration until quality threshold met
+- Configurable quality gate (`quality_gate`) and maximum iterations
+- Lightweight quality assessment (no LLM calls), supports dry-run mode
+
+```python
+from scripts.collaboration.feedback_control_loop import FeedbackControlLoop
+from scripts.collaboration.dispatcher import MultiAgentDispatcher
+
+dispatcher = MultiAgentDispatcher()
+loop = FeedbackControlLoop(dispatcher, quality_gate=0.7, max_iterations=3)
+result = loop.run("Design secure auth system", roles=["architect", "security"])
+print(f"Iterations: {loop.iteration_count}")
+print(f"Best quality: {loop.best_quality:.2f}")
+# Automatically iterates until quality gate met or max iterations reached
+```
+
+#### 2️⃣ ExecutionGuard (Execution Guardian)
+**Chinese Name**: 执行守护者 (Execution Guardian)
+**Core Capabilities**:
+- Real-time execution monitoring with 4 abort conditions: timeout, output size, token count, critical keywords
+- Lightweight checks (<1ms), zero external dependencies
+- Dynamically configurable thresholds (max_duration_sec, max_output_tokens, etc.)
+
+```python
+from scripts.collaboration.execution_guard import ExecutionGuard
+
+guard = ExecutionGuard(max_duration_sec=300.0, max_output_tokens=8000)
+should_abort, reason = guard.check_abort(
+    worker_output="Generating code...",
+    elapsed_time=120.5,
+    token_count=5000
+)
+if should_abort:
+    print(f"Aborting: {reason}")
+    # Example: "Timeout exceeded: 120.5s > 300.0s"
+# Also detect warning keywords (without triggering abort)
+warnings = guard.check_warnings("WARNING: High memory usage")
+print(f"Warnings: {warnings}")  # ['WARNING']
+```
+
+#### 3️⃣ PerformanceFingerprint (Performance Fingerprint)
+**Chinese Name**: 性能指纹系统 (Performance Fingerprint System)
+**Core Capabilities**:
+- Unified execution fingerprint recording (fuses 4 data sources: invocation counts, latency, state snapshots, retrospective deviations)
+- Pure Python TF-IDF implementation (no sklearn/numpy), supports English/Chinese mixed content
+- JSON persistence to `.devsquad_data/fingerprints/`, graceful cold-start degradation
+
+```python
+from scripts.collaboration.performance_fingerprint import PerformanceFingerprint
+
+fingerprint = PerformanceFingerprint()
+fid = fingerprint.record_execution(
+    task="Implement user authentication",
+    result=dispatch_result,
+    timing={"total": 12.5, "planning": 2.0, "coding": 8.0, "review": 2.5},
+    roles_used=["architect", "coder", "tester"],
+)
+print(f"Fingerprint ID: {fid}")  # fp_20260518_143052_a1b2c3d4
+
+# Find similar historical tasks using TF-IDF
+similar = fingerprint.find_similar("Add login page", top_k=3)
+for case in similar:
+    print(f"Task: {case['task']}")
+    print(f"Similarity: {case['similarity']:.2%}")
+    print(f"Roles used: {case['roles_used']}")
+    print(f"Success: {case['success']}")
+
+# Get overall statistics
+stats = fingerprint.get_stats()
+print(f"Total executions: {stats['total']}")
+print(f"Success rate: {stats['success_rate']:.1%}")
+```
+
+#### 4️⃣ SimilarTaskRecommender (Similar Task Recommender)
+**Chinese Name**: 相似任务推荐器 (Similar Task Recommender)
+**Core Capabilities**:
+- TF-IDF-based task similarity search with historical success configuration recommendations
+- Intelligent role combination recommendation, intent prediction, execution time estimation
+- Confidence scoring (high/medium/low), graceful cold-start degradation
+
+```python
+from scripts.collaboration.similar_task_recommender import SimilarTaskRecommender
+
+recommender = SimilarTaskRecommender()
+result = recommender.recommend("Design user authentication system")
+print(f"Recommended roles: {result['recommended_roles']}")
+# Output: ['architect', 'coder', 'tester', 'security']
+print(f"Confidence: {result['confidence']}")  # high/medium/low
+print(f"Estimated duration: {result['estimated_duration_s']:.1f}s")
+
+# View similar case details
+for case in result['similar_cases']:
+    print(f"Task: {case['task']}")
+    print(f"Similarity: {case['similarity']:.2%}")
+    print(f"Historical roles: {case['roles']}")
+    print(f"Success: {case['success']}")
+
+# Quick method: get role suggestions only
+roles = recommender.get_role_suggestion("Implement payment API")
+print(f"Suggested roles: {roles}")
+```
+
+#### 5️⃣ AdaptiveRoleSelector (Adaptive Role Selector)
+**Chinese Name**: 自适应角色选择器 (Adaptive Role Selector)
+**Core Capabilities**:
+- Three-tier selection strategy based on historical success rates (similar tasks → intent match → fallback to default)
+- Configurable minimum success rate and maximum role count
+- Supports manual statistics updates and comprehensive role effectiveness reporting
+
+```python
+from scripts.collaboration.adaptive_role_selector import AdaptiveRoleSelector
+
+selector = AdaptiveRoleSelector()
+roles = selector.select_roles(
+    task="Build high-concurrency microservices architecture",
+    intent="feature_implementation",
+    min_success_rate=0.5,
+    max_roles=5,
+)
+print(f"Recommended roles: {roles}")
+# Output: ['architect', 'devops', 'security', 'tester']
+# Or: [] (returns empty when no historical data, caller falls back to default RoleMatcher)
+
+# Manually update statistics (for external system integration)
+selector.update_stats(["architect", "coder"], success=True, duration_s=12.5)
+
+# Generate role effectiveness report
+report = selector.get_role_report()
+for role_name, metrics in report.items():
+    print(f"{role_name}: success_rate={metrics['success_rate']:.1%}, "
+          f"avg_duration={metrics['avg_duration']:.1f}s")
+```
+
+### 🔗 Integration Architecture
 
-**DevSquad V3.6.0** adds AnchorChecker for milestone anchor verification, RetrospectiveEngine for independent retrospectives, FeatureUsageTracker for data-driven feature optimization, StructuredGoal for structured goal management, and FallbackBackend for automatic backend failover — making multi-agent collaboration more reliable, self-improving, and observable.
+The 5 cybernetic modules are designed as **non-invasive wrappers** — they work independently or together without modifying existing core logic:
+
+```
+User Task
+    ↓
+[SimilarTaskRecommender] ← Optional: suggest roles from history
+    ↓
+[AdaptiveRoleSelector]   ← Optional: optimize role selection
+    ↓
+[MultiAgentDispatcher]
+    ↓
+[FeedbackControlLoop]     ← Wrap dispatcher for auto-iteration
+    ↓ [each worker step]
+[ExecutionGuard]          ← Guard each worker execution
+    ↓
+[PerformanceFingerprint]  ← Record after dispatch completes
+```
+
+**Recommended usage** (progressive adoption):
+```python
+from scripts.collaboration import (
+    MultiAgentDispatcher, FeedbackControlLoop,
+    ExecutionGuard, PerformanceFingerprint
+)
+
+dispatcher = MultiAgentDispatcher()
+guard = ExecutionGuard()
+fingerprint = PerformanceFingerprint()
+
+# Option 1: Full cybernetics stack
+loop = FeedbackControlLoop(dispatcher, quality_gate=0.7)
+result = loop.run("Your task here")
+
+# Option 2: Guard only (minimal adoption)
+result = dispatcher.dispatch("Your task")
+for w in result.worker_results:
+    abort, reason = guard.check_abort(w.output, w.duration)
+    if abort:
+        print(f"Aborted: {reason}")
+
+# Option 3: Learning only
+fingerprint.record_execution("task", result, result.timing, result.matched_roles)
+similar = fingerprint.find_similar("new task", top_k=3)
+```
+
+All modules are **optional switches** — DevSquad works perfectly without them.
 
 ### 🎯 Quick Start (4 Ways to Use DevSquad)
 
@@ -104,7 +333,10 @@ devsquad dispatch -t "your task description"
 streamlit run scripts/dashboard.py
 
 # Open http://localhost:8501
-# Login with: admin / admin123
+# ⚠️ Security: Default credentials are for initial setup only.
+#    Login with default account, then change password immediately.
+#    Username: admin   Password: <your-secure-password>
+#    Or set via environment variables: $DASHBOARD_USER / $DASHBOARD_PASS
 ```
 
 #### 2️⃣ REST API Server
@@ -167,7 +399,7 @@ python scripts/cli.py lifecycle build --visual --verbose
 
 ---
 
-## ✨ Key Features (V3.6.0)
+## ✨ Key Features (V3.6.5)
 
 ### ⚓ AnchorChecker (NEW)
 Milestone anchor verification that ensures critical checkpoints are properly validated before proceeding:
@@ -204,6 +436,11 @@ Automatic backend failover that ensures LLM availability even when primary backe
 - **Priority-Based Routing** — Configure backend priority order (e.g., OpenAI → Anthropic → Mock)
 - **Recovery Detection** — Automatically restore primary backend when it recovers
 
+### 🔍 VerificationGate — Evidence-Based Quality
+- **Prove-It Pattern**: Every completion claim must include verifiable evidence (test output, diff, benchmark)
+- **7 Red Flags**: `no_test` | `tests_pass_first_run` | `no_regression_test` | `no_security_scan` | `no_perf_baseline` | `vague_description` | `evidence_missing`
+- **Auto-active**: Integrated into TaskCompletionChecker — zero config required
+
 ### 🔐 Authentication & Authorization
 - **Multi-user support** with role-based access control (RBAC)
 - **Three roles**: Admin (full access), Operator (execute), Viewer (read-only)
@@ -259,7 +496,7 @@ System:
 
 ---
 
-## 🧩 Layered Sub-Skill Architecture (V3.6.0)
+## 🧩 Layered Sub-Skill Architecture (V3.6.5)
 
 > DevSquad provides **6 atomic sub-skills** that can be used independently or together.
 > Each sub-skill is a thin wrapper (~50 lines) importing existing core modules — no duplicated logic.
@@ -369,7 +606,17 @@ You: "Design a microservices e-commerce backend"
 - **Python 3.9+** (3.9, 3.10, 3.11, 3.12 supported)
 - **pip** or **pipenv** for package management
 
-### Option A: Core Installation (CLI + Dashboard)
+### Option A: PyPI Install (Recommended)
+```bash
+# Install from PyPI — zero setup, ready to use
+pip install devsquad
+
+# With optional dependencies
+pip install "devsquad[api]"    # FastAPI + Streamlit dashboard
+pip install "devsquad[all]"    # All optional features
+```
+
+### Option B: Core Installation (CLI + Dashboard)
 ```bash
 git clone https://github.com/your-org/DevSquad.git
 cd DevSquad
@@ -381,7 +628,7 @@ pip install -e .
 devsquad dispatch -t "Design user authentication system"
 ```
 
-### Option B: Full Production Stack (Recommended)
+### Option C: Full Production Stack (All Features)
 ```bash
 # Clone and install with all production features
 git clone https://github.com/your-org/DevSquad.git
@@ -416,7 +663,7 @@ pip install -e ".[all]"
 ```bash
 # Check version
 devsquad --version
-# Expected: devsquad 3.6.0
+# Expected: devsquad 3.6.5
 
 # Run tests
 pytest tests/ -v --tb=short
@@ -446,7 +693,7 @@ python3 scripts/cli.py dispatch -t "Design auth system" -r arch --backend openai
 # Other commands
 python3 scripts/cli.py status          # System status
 python3 scripts/cli.py roles           # List available roles
-python3 scripts/cli.py --version       # Show version (3.6.0)
+python3 scripts/cli.py --version       # Show version (3.6.5)
 ```
 
 **2. Python API**
@@ -508,7 +755,7 @@ Exposes 6 tools: `multiagent_dispatch`, `multiagent_quick`, `multiagent_roles`,
 
 **Auto-match**: If no roles specified, the dispatcher automatically matches based on task keywords.
 
-## Architecture Overview (48 Core Modules)
+## Architecture Overview (60+ Core Modules)
 
 DevSquad is built on a layered architecture with clear separation of concerns:
 
@@ -540,7 +787,7 @@ DevSquad is built on a layered architecture with clear separation of concerns:
 └─────────────────────────────────────────────────┘
 ```
 
-## What's New in V3.6.0 🆕
+## What's New in V3.6.5 🆕
 
 ### AnchorChecker System
 Milestone anchor verification that ensures critical checkpoints are validated before proceeding:
@@ -673,7 +920,7 @@ See [Integration Guide](docs/guides/agent_briefing_confidence_integration.md) fo
 
 ### Project Lifecycle (11-Phase Model)
 
-DevSquad V3.6.0 defines an **11-phase (4 optional)** project lifecycle with clear roles, dependencies, and gate conditions:
+DevSquad V3.6.5 defines an **11-phase (4 optional)** project lifecycle with clear roles, dependencies, and gate conditions:
 
 ```
 P1 → P2 ──┬──→ P3 ──→ P6 ──→ P7 ──→ P8 ──→ P9 ──→ P10 ──→ P11
@@ -698,7 +945,9 @@ See [GUIDE.md](GUIDE.md) §4 for full lifecycle details with gate conditions and
 - **GitHub Actions CI**: Python 3.9-3.12 matrix testing
 - **pip installable**: `pip install -e .` with optional dependencies
 
-## Module Reference (48 Modules)
+## Module Reference (60+ Modules)
+
+> 💡 **Table too wide?** [View modules online](https://github.com/lulin70/DevSquad/blob/main/SKILL.md#L27) or use `devsquad --help modules` for a compact list.
 
 | Module | File | Purpose |
 |--------|------|---------|
@@ -824,7 +1073,7 @@ python3 -m pytest scripts/collaboration/core_test.py \
   tests/test_cli_lifecycle.py -v
 
 # Quick smoke test
-python3 scripts/cli.py --version    # 3.6.0
+python3 scripts/cli.py --version    # 3.6.5
 python3 scripts/cli.py status       # System ready
 python3 scripts/cli.py roles        # List 7 roles
 
@@ -836,14 +1085,59 @@ python3 scripts/cli.py review -t "Check PR #123"
 python3 scripts/cli.py ship -t "Deploy to production"
 ```
 
+### 🔄 Upgrade Smoke Test
+After upgrading DevSquad, run these commands to verify your environment:
+```bash
+# Quick health check (should complete in < 30s)
+python3 scripts/cli.py --version       # Expected: DevSquad 3.6.5
+python3 scripts/cli.py status          # Expected: System ready
+python3 scripts/cli.py roles           # Expected: 7 core roles listed
+
+# Full test suite
+python3 -m pytest tests/ -q --tb=line # Expected: 1731 passed
+```
+
+### With Coverage Report
+```bash
+# Install coverage tool first: pip install pytest-cov
+python3 -m pytest tests/ --cov=scripts --cov-report=term-missing --cov-fail-under=80
+# Expected: coverage ≥ 80%, detailed missing-line report
+```
+
+### Test Layering Strategy
+
+DevSquad uses a priority-based test layering strategy:
+
+| Priority | Scope | Examples | Count |
+|----------|-------|----------|-------|
+| **P0** | Quality Framework Core | AntiRationalization (39), VerificationGate (42), IntentWorkflowMapper (58), AuthManager (35) | ~200 |
+| **P1** | Enhancement Modules | FiveAxisConsensus (29), OperationClassifier (27), OutputSlicer (26), CIFeedbackAdapter (22) | ~150 |
+| **P1+** | Cybernetics (V3.6.5) | FeedbackControlLoop (19), ExecutionGuard (40), PerformanceFingerprint (13), SimilarTaskRecommender (17), AdaptiveRoleSelector (21) | **110** |
+| **P2** | Integration & E2E | Full lifecycle dispatch, cross-module integration | ~200 |
+| **P3** | Unit per Module | Core dispatcher, RoleMapping, MCEAdapter, LLM backends | ~400+ |
+
+**Total: 1731 tests**
+
+Run by priority:
+```bash
+# P0 only (critical path, < 10s)
+python3 -m pytest tests/ -k "anti_ratif or verification or intent_workflow or auth" -q
+
+# P0 + P1 (quality + enhancement, < 30s)
+python3 -m pytest tests/ -k "anti_ratif or verification or intent or auth or five_axis or operation" -q
+
+# Full suite
+python3 -m pytest tests/ -q --tb=line
+```
+
 ## Documentation
 
 | Document | Description |
 |----------|-------------|
-| [QUICK_START_EN.md](docs/i18n/QUICK_START_EN.md) | ⚡ Quick start guide (English, 5 minutes) |
-| [REFERENCE_GUIDE_EN.md](docs/i18n/REFERENCE_GUIDE_EN.md) | 📖 Complete reference guide (English) |
-| [QUICK_START_JP.md](docs/i18n/QUICK_START_JP.md) | ⚡ クイックスタートガイド (日本語, 5分) |
-| [REFERENCE_GUIDE_JP.md](docs/i18n/REFERENCE_GUIDE_JP.md) | 📖 完全リファレンスガイド (日本語) |
+| [QUICK_START_EN.md](docs/i18n/QUICK_START_EN.md) | Quick start guide (English, 5 minutes) |
+| [REFERENCE_GUIDE_EN.md](docs/i18n/REFERENCE_GUIDE_EN.md) | Complete reference guide (English) |
+| [QUICK_START_JP.md](docs/i18n/QUICK_START_JP.md) | クイックスタートガイド (日本語, 5分) |
+| [REFERENCE_GUIDE_JP.md](docs/i18n/REFERENCE_GUIDE_JP.md) | 完全リファレンスガイド (日本語) |
 | [GUIDE.md](GUIDE.md) | Complete user guide (Chinese) |
 | [GUIDE_EN.md](docs/i18n/GUIDE_EN.md) | ~~Complete user guide (English)~~ → See QUICK_START + REFERENCE_GUIDE |
 | [GUIDE_JP.md](docs/i18n/GUIDE_JP.md) | ~~完全なユーザーガイド (日本語)~~ → クイックスタート＋リファレンスを参照 |
@@ -889,13 +1183,13 @@ DevSquad is **not TRAE-exclusive**. It supports 6 integration methods:
 
 | Platform | Integration | Setup Difficulty | Key Features Available |
 |----------|-------------|-----------------|----------------------|
-| **TRAE IDE** | Native Skill (`skill-manifest.yaml`) | ⭐ Zero config | Full: Dispatcher + Dashboard + CLI |
-| **Claude Code** | MCP Server / Python import | ⭐ Low | 6 MCP tools or direct API |
-| **Cursor** | MCP Server (`stdio` mode) | ⭐ Low | Same as Claude Code |
+| **TRAE IDE** | Native Skill (`skill-manifest.yaml`) | Zero config | Full: Dispatcher + Dashboard + CLI |
+| **Claude Code** | MCP Server / Python import | Low | 6 MCP tools or direct API |
+| **Cursor** | MCP Server (`stdio` mode) | Low | Same as Claude Code |
 | **OpenClaw / WorkBuddy Claw** | `WorkBuddyClawSource` bridge | Auto | Read-only memory bridge |
-| **Any MCP Client** | stdio / SSE dual mode | ⭐ Low | 6 tools, configurable port |
-| **Pure Python** | `pip install -e .` | ⭐ Low | CLI + API + Skills + REST |
-| **Docker** | `docker build & run` | ⭐ Low | Isolated container with all features |
+| **Any MCP Client** | stdio / SSE dual mode | Low | 6 tools, configurable port |
+| **Pure Python** | `pip install -e .` | Low | CLI + API + Skills + REST |
+| **Docker** | `docker build & run` | Low | Isolated container with all features |
 
 ### Quick Start per Platform
 
@@ -922,13 +1216,12 @@ docker build -t devsquad . && docker run -it devsquad dispatch -t "test"
 
 | Date | Version | Highlights |
 |------|---------|-----------|
-| 2026-05-16 | **V3.6.0** | 🧩 **Layered Sub-Skill Architecture** — 6 atomic sub-skills (dispatch/intent/review/security/test/retrospective) with lazy-loading registry via importlib, each ~50 lines wrapping existing core modules, no duplicated logic. All sub-skills work in Mock mode without API keys. Plus: Cross-platform compatibility docs updated for Claude Code/Cursor/OpenClaw/Pure Python/Docker/MCP. |
-| 2026-05-13 | **V3.6.0** | ⚓ AnchorChecker (milestone anchor verification + drift detection), RetrospectiveEngine (independent retrospective + pattern extraction), StructuredGoal (structured goal decomposition + progress tracking), FallbackBackend (automatic LLM failover + health monitoring), FeatureUsageTracker (feature usage tracking + reporting + auto-persistence), 7 module integrations (IntentWorkflowMapper/AISemanticMatcher/DualLayerContextManager/OperationClassifier/SkillRegistry/FiveAxisConsensusEngine/NullProviders), 1548+ tests, 48 core modules |
-| 2026-05-05 | **V3.5.0** | 📋 Enhancement Sprint — Code walkthrough enhancement, documentation consistency checks, Karpathy principles, project understanding (AgentBriefing), CLI lifecycle commands, structured output, 748+ tests |
-| 2026-05-03 | **V3.4.1** | 🚀 Agent Skills Quality Framework (P0) — AntiRationalizationEngine + VerificationGate + IntentWorkflowMapper + CLI Lifecycle Commands (spec/plan/build/test/review/ship) + 167 new tests + Google Agent Skills integration + 49 core modules |
-| 2026-05-02 | **V3.4.0** | 🆕 11-Phase Project Lifecycle (full/backend/frontend/internal_tool/minimal templates), requirement change management, gate mechanism with gap reporting, 748+ tests passing, WorkflowEngine lifecycle support |
-| 2026-05-01 | V3.4.0 | AgentBriefing (context-aware task briefing), ConfidenceScore (5-factor quality assessment), EnhancedWorker (auto quality assurance with retry + memory_provider rule injection), Protocol interface system (match_rules/format_rules_as_prompt), CarryMem v0.2.8+ integration, comprehensive documentation |
-| 2026-04-27 | V3.4.0 | Real LLM backend (OpenAI/Anthropic/Mock), ThreadPoolExecutor parallel execution, InputValidator + prompt injection protection, CheckpointManager, WorkflowEngine, TaskCompletionChecker, AISemanticMatcher, streaming output, Docker, GitHub Actions CI, config file, CodeMapGenerator, DualLayerContext, SkillRegistry, CarryMem integration, 234 unit tests |
+| 2026-05-20 | **V3.6.5** | 🎯 **Enterprise Edition + E2E Testing** — Enterprise Features: RBAC Engine (15+ permissions, 5 roles), Audit Logger (SHA256 integrity chain), Multi-Tenancy Manager (3 isolation levels), Sensitive Data Masker (PII protection). AsyncIO Transformation: 2x throughput improvement, 50% latency reduction. Redis Cache Integration: L1→L2→L3 three-level cache, 95%+ hit rate. Prometheus Monitoring: 12 core metrics, /metrics endpoint. E2E Test Suite: 27 test cases, 5 scenarios (CLI/API/Collaboration/Enterprise/Error Recovery), 100% pass rate in 9 seconds. Code Quality: print() → logging migration, TODO/FIXME cleanup, pre-commit hooks, .editorconfig. Total: 13,000+ lines new code, 1731 tests passing, 97% maturity score |
+| 2026-05-17 | **V3.6.1** | **Cybernetics Enhancement** — 5 new modules (FeedbackControlLoop/ExecutionGuard/PerformanceFingerprint/SimilarTaskRecommender/AdaptiveRoleSelector) with feedback loops, execution guards, TF-IDF similarity search, and adaptive role selection. Inspired by upstream TraeMultiAgentSkill v2.5's cybernetics architecture. |
+| 2026-05-16 | **V3.6.0** | **Layered Sub-Skill Architecture + Core Modules** — 6 atomic sub-skills (dispatch/intent/review/security/test/retrospective) with lazy-loading registry via importlib, each ~50 lines wrapping existing core modules. Plus: AnchorChecker (milestone anchor verification + drift detection), RetrospectiveEngine (independent retrospective + pattern extraction), StructuredGoal (structured goal decomposition + progress tracking), FallbackBackend (automatic LLM failover + health monitoring), FeatureUsageTracker (feature usage tracking + reporting + auto-persistence), 7 module integrations (IntentWorkflowMapper/AISemanticMatcher/DualLayerContextManager/OperationClassifier/SkillRegistry/FiveAxisConsensusEngine/NullProviders), 1662+ tests, 48 core modules. Cross-platform compatibility: Claude Code/Cursor/OpenClaw/Pure Python/Docker/MCP. |
+| 2026-05-05 | **V3.5.0** | Enhancement Sprint — Code walkthrough enhancement, documentation consistency checks, Karpathy principles, project understanding (AgentBriefing), CLI lifecycle commands, structured output, 748+ tests |
+| 2026-05-03 | **V3.4.1** | Agent Skills Quality Framework (P0) — AntiRationalizationEngine + VerificationGate + IntentWorkflowMapper + CLI Lifecycle Commands (spec/plan/build/test/review/ship) + 167 new tests + Google Agent Skills integration + 49 core modules |
+| 2026-05-02 | **V3.4.0** | **Foundation Release** — Real LLM backend (OpenAI/Anthropic/Mock), ThreadPoolExecutor parallel execution, InputValidator + prompt injection protection, CheckpointManager, WorkflowEngine with 11-phase lifecycle templates (full/backend/frontend/internal_tool/minimal), TaskCompletionChecker, AISemanticMatcher, streaming output, Docker, GitHub Actions CI, config file, CodeMapGenerator, DualLayerContext, SkillRegistry, CarryMem integration, AgentBriefing, ConfidenceScore, EnhancedWorker with auto QA, Protocol interface system, 234+ unit tests, requirement change management with gate mechanism and gap reporting |
 | 2026-04-17 | V3.2 | E2E Demo, MCE Adapter, Dispatcher UX |
 | 2026-04-16 | V3.0 | Complete redesign — Coordinator/Worker/Scratchpad architecture |