ADR	Title	Status	File
ADR-001	Adopt agentic-flow as Core Foundation	✅ Implemented	ADR-001-AGENT-IMPLEMENTATION.md
ADR-002	Implement Domain-Driven Design Structure	✅ Implemented	ADR-002-DDD-STRUCTURE.md
ADR-003	Single Coordination Engine	✅ Implemented	ADR-003-CONSOLIDATION-COMPLETE.md
ADR-004	Plugin-Based Architecture	✅ Implemented	ADR-004-PLUGIN-ARCHITECTURE.md
ADR-005	MCP-First API Design	✅ Implemented	ADR-005-implementation-summary.md
ADR-006	Unified Memory Service	✅ Implemented	ADR-006-UNIFIED-MEMORY.md
ADR-007	Event Sourcing for State Changes	✅ Implemented	ADR-007-EVENT-SOURCING.md
ADR-008	Vitest Over Jest	✅ Implemented	ADR-008-VITEST.md
ADR-009	Hybrid Memory Backend as Default	✅ Implemented	ADR-009-IMPLEMENTATION.md
ADR-010	Remove Deno Support	✅ Implemented	ADR-010-NODE-ONLY.md

Providers & Security (ADR-011 to ADR-017)

ADR	Title	Status	File
ADR-011	LLM Provider System	✅ Implemented	ADR-011-llm-provider-system.md
ADR-012	MCP Security Features	✅ Implemented	ADR-012-mcp-security-features.md
ADR-013	Core Security Module	✅ Implemented	ADR-013-core-security-module.md
ADR-014	Workers System	✅ Implemented	ADR-014-workers-system.md
ADR-015	Unified Plugin System	✅ Implemented	ADR-015-unified-plugin-system.md
ADR-016	Collaborative Issue Claims	✅ Implemented	ADR-016-collaborative-issue-claims.md
ADR-017	RuVector Integration	✅ Implemented	ADR-017-ruvector-integration.md

Advanced Features (ADR-018 to ADR-025)

ADR	Title	Status	File
ADR-018	Claude Code Integration	✅ Implemented	ADR-018-claude-code-integration.md
ADR-019	Headless Runtime Package	✅ Implemented	ADR-019-headless-runtime-package.md
ADR-020	Headless Worker Integration	✅ Implemented	ADR-020-headless-worker-integration.md
ADR-021	Transfer Hook IPFS Pattern Sharing	✅ Implemented	ADR-021-transfer-hook-ipfs-pattern-sharing.md
ADR-022	AIDefence Integration	✅ Implemented	ADR-022-aidefence-integration.md
ADR-023	ONNX Hyperbolic Embeddings Init	✅ Implemented	ADR-023-onnx-hyperbolic-embeddings-init.md
ADR-024	Embeddings MCP Tools	✅ Implemented	ADR-024-embeddings-mcp-tools.md
ADR-025	Auto-Update System	✅ Implemented	ADR-025-auto-update-system.md

Neural & Integration (ADR-026 to ADR-031)

ADR	Title	Status	File
ADR-026	Agent Booster Model Routing	✅ Implemented	ADR-026-agent-booster-model-routing.md
ADR-027	RuVector PostgreSQL Integration	✅ Implemented	ADR-027-ruvector-postgresql-integration.md
ADR-028	Neural Attention Mechanisms	✅ Implemented	ADR-028-neural-attention-mechanisms.md
ADR-029	GNN Integration	✅ Implemented	ADR-029-gnn-integration.md
ADR-030	Agentic QE Integration	✅ Implemented	ADR-030-agentic-qe-integration.md
ADR-031	Prime Radiant Integration	✅ Implemented	ADR-031-prime-radiant-integration.md

Packaging & Branding (ADR-042 to ADR-048)

ADR	Title	Status	File
ADR-046	Dual Umbrella: claude-flow + ruflo	Accepted	ADR-046-ruflo-rebrand.md
ADR-047	Fast Mode Integration	Proposed	ADR-047-fast-mode-integration.md
ADR-048	Auto Memory Integration	Accepted	ADR-048-auto-memory-integration.md

RuVector WASM Plugin Architecture (ADR-032 to ADR-041)

ADR	Title	Category	Status	File
ADR-032	Healthcare Clinical Decision Support	Practical	Proposed	ADR-032-healthcare-clinical-plugin.md
ADR-033	Financial Risk Analysis	Practical	Proposed	ADR-033-financial-risk-plugin.md
ADR-034	Legal Contract Analysis	Practical	Proposed	ADR-034-legal-contract-plugin.md
ADR-035	Advanced Code Intelligence	Advanced	Proposed	ADR-035-code-intelligence-plugin.md
ADR-036	Test Intelligence	Advanced	Proposed	ADR-036-test-intelligence-plugin.md
ADR-037	Performance Optimization	Advanced	Proposed	ADR-037-performance-optimization-plugin.md
ADR-038	Multi-Agent Neural Coordination	Cutting-Edge	Proposed	ADR-038-multi-agent-coordination-plugin.md
ADR-039	Cognitive Kernel	Cutting-Edge	Proposed	ADR-039-cognitive-kernel-plugin.md
ADR-040	Quantum-Inspired Optimization	Exotic SOTA	Proposed	ADR-040-quantum-inspired-plugin.md
ADR-041	Hyperbolic Reasoning	Exotic SOTA	Proposed	ADR-041-hyperbolic-reasoning-plugin.md

Implementation Progress

Last Updated: 2026-01-13 (Beta Ready Audit) Status: ✅ READY FOR 3.0.0-beta.1

Component	Status	Details
DDD Modules	✅ 100%	16 packages, 419 TS files
CLI Commands	✅ 100%	28 commands, 140+ subcommands
MCP Tools	✅ 100%	119 tools (exceeds V2)
Hooks System	✅ 100%	20 CLI + 60 MCP hook tools
Test Coverage	✅ 85%+	85+ test files (ADR-008)
Service Integration	✅ Complete	agentic-flow@alpha integration
RuVector Integration	✅ Complete	Q-Learning, AST, Diff, Coverage (ADR-017)
Performance Benchmarks	✅ Complete	Full benchmark suite
Auto-Update System	✅ Complete	Rate-limited startup checks (ADR-025)
Production Hardening	✅ Complete	Real metrics, labeled examples, fallback warnings

Overall V3 Implementation: 100% Complete

Beta Readiness Checklist

Category	Status
Real ONNX embeddings	✅
Real performance metrics	✅
Real security scanning	✅
Fallback warnings	✅
Auto-update system	✅
Claims MCP tools	✅
Production hardening	✅
Windows validated	✅

Quick Summary

Core Decisions

ADR-001: Build on agentic-flow@alpha instead of duplicating (eliminates 10,000+ lines)
ADR-002: Domain-Driven Design with bounded contexts for clean architecture
ADR-003: Single UnifiedSwarmCoordinator as canonical coordination engine
ADR-004: Microkernel with plugins for optional features (HiveMind, Neural, etc.)
ADR-005: MCP tools as primary API, CLI as thin wrapper

Technical Decisions

ADR-006: Single MemoryService with SQLite, AgentDB, or Hybrid backends
ADR-007: Event sourcing for audit trail and state reconstruction
ADR-008: Vitest for 10x faster testing with native ESM
ADR-009: Hybrid backend (SQLite + AgentDB) as default for best performance
ADR-010: Node.js 20+ only, removing Deno complexity

Additional Files

v3-adrs.md - Complete ADR master document with all decisions
ADR-003-implementation-status.md - Detailed implementation tracking

Performance Targets (from ADRs)

Metric	Target	ADR Reference
Code reduction	<5,000 lines vs 15,000+	ADR-001
HNSW search	150x-12,500x faster	ADR-009
Flash Attention	2.49x-7.47x speedup	ADR-001
Test execution	<5s (10x improvement)	ADR-008
Startup time	<500ms	ADR-004
Query latency	<100ms	ADR-006

Security Improvements

All ADRs consider security:

CVE-1: Command injection prevention (ADR-005 input validation)
CVE-2: Path traversal prevention (ADR-006 memory sandboxing)
CVE-3: Credential generation (secure random with rejection sampling)

Last Updated: 2026-01-13 Project: Claude-Flow V3 Version: 3.0.0-alpha.84 (Beta Ready)

Recent Updates (2026-01-13)

Release: @claude-flow/cli@3.0.0-alpha.84 (Beta Ready)

All Audit Issues Resolved:

Fix	Before	After
Profile metrics	Hardcoded 23%, 145MB	Real: process.memoryUsage(), process.cpuUsage()
CVE data	Unmarked fake data	Labeled as examples with warnings
Demo mode warnings	Silent fallback	⚠ DEMO MODE / OFFLINE MODE warnings

Auto-Update System (ADR-025):

npx claude-flow update check      # Check for updates
npx claude-flow update all        # Update all packages
npx claude-flow update history    # View update history
npx claude-flow update rollback   # Rollback last update

Previous Updates (2026-01-07)

Release: @claude-flow/cli@3.0.0-alpha.15 (Latest)

Doctor Command Enhancements:

Claude Code CLI Check: Verifies @anthropic-ai/claude-code installation
Auto-Install: --install flag to auto-install missing Claude Code CLI
Fixed Package Paths: Corrected dist/src/ paths for proper npm resolution

# Check system health including Claude Code CLI
npx claude-flow@v3alpha doctor

# Auto-install Claude Code CLI if missing
npx claude-flow@v3alpha doctor --install

# Check only Claude Code CLI
npx claude-flow@v3alpha doctor -c claude

Package Resolution Fix: Fixed Windows module resolution issue where @claude-flow/cli exports pointed to wrong paths (dist/index.js → dist/src/index.js).

Release: @claude-flow/cli@3.0.0-alpha.7

Hive-Mind CLI: All MCP tools now exposed via CLI subcommands:
- hive-mind join <agent-id> - Join agent to hive
- hive-mind leave <agent-id> - Remove agent from hive
- hive-mind consensus - Manage consensus proposals and voting
- hive-mind broadcast -m <msg> - Broadcast messages to workers
- hive-mind memory - Access shared memory (get/set/delete/list)
Bug Fix: Fixed positional argument parsing for subcommands in CLI parser
File Persistence: All MCP tools use file-based persistence in .claude-flow/ directories
ADR-014: Node.js Worker Daemon - cross-platform TypeScript daemon replaces shell helpers
CLI: daemon command with start/stop/status/trigger/enable subcommands
Session Integration: Auto-start daemon on SessionStart, auto-stop on SessionEnd

CLI MCP Tool Coverage

Category	Tools	CLI Status
Agent	spawn, terminate, status, list, pool, health, update	✅ Complete
Hive-Mind	init, spawn, status, task, join, leave, consensus, broadcast, memory, optimize-memory, shutdown	✅ Complete
Task	create, status, list, complete, cancel	✅ Complete
Session	save, restore, list, delete, export	✅ Complete
Config	get, set, list, reset, export, import	✅ Complete
Memory	store, retrieve, list, delete, search	✅ Complete
Workflow	create, execute, list, status, delete	✅ Complete

Install

npx @claude-flow/cli@v3alpha --help

Release: @claude-flow/cli@3.0.0-alpha.11 (2026-01-07)

New V3 Advanced CLI Commands

All commands include subcommand help and "Created with ❤️ by ruv.io" branding.

Command	Description	Subcommands
`neural`	Neural pattern training, MoE, Flash Attention	train, status, patterns, predict, optimize
`security`	Security scanning, CVE detection, threat modeling	scan, cve, threats, audit, secrets
`performance`	Performance profiling, benchmarking, optimization	benchmark, profile, metrics, optimize, bottleneck
`providers`	AI provider management, models, configurations	list, configure, test, models, usage
`plugins`	Plugin management, installation, lifecycle	list, install, uninstall, toggle, info, create
`deployment`	Deployment management, environments, rollbacks	deploy, status, rollback, history, environments, logs
`claims`	Claims-based authorization, access control	list, check, grant, revoke, roles, policies
`embeddings`	Vector embeddings, semantic search	generate, search, compare, collections, index, providers
`doctor`	System diagnostics	Node version, config, daemon, memory, API keys, MCP, disk
`completions`	Shell completions	bash, zsh, fish, powershell

P0 Features Completed (alpha.12)

Smart Error Suggestions: Typo detection with Levenshtein distance

$ claude-flow swram
[ERROR] Unknown command: swram
  Did you mean one of these?
  - swarm
  - neural
  - start

$ claude-flow memroy
[ERROR] Unknown command: memroy
  Did you mean "memory"?

Doctor Command: System health diagnostics

$ claude-flow doctor
Claude Flow Doctor
──────────────────────────────────────────────────
✓ Node.js Version: v22.21.1 (>= 20 required)
✓ npm Version: v10.9.4
✓ Git: v2.52.0
✓ Git Repository: In a git repository
⚠ Config File: No config file (using defaults)
⚠ Daemon Status: Not running
⚠ Memory Database: Not initialized
⚠ API Keys: No API keys found
⚠ MCP Servers: No MCP config found
✓ Disk Space: 73G available
✓ TypeScript: v5.9.3
──────────────────────────────────────────────────
Summary: 6 passed, 5 warnings

Shell Completions: Tab completion for all shells

# Install bash completions
claude-flow completions bash > ~/.bash_completion.d/claude-flow

# Install zsh completions
claude-flow completions zsh > ~/.zfunc/_claude-flow

# Install fish completions
claude-flow completions fish > ~/.config/fish/completions/claude-flow.fish

# Install PowerShell completions
claude-flow completions powershell >> $PROFILE

CLI Roadmap

Priority Recommendations

Priority	Recommendation	Status	Description
🔴 P0	Add `doctor` command	✅ Done	System diagnostics, dependency checks, config validation
🔴 P0	Add `completions` command	✅ Done	Shell completions for bash, zsh, fish, powershell
🔴 P0	Add smart error suggestions	✅ Done	Levenshtein distance for typo corrections
🟡 P1	Resolve provider config overlap	Pending	Unify provider configs across embeddings/providers commands
🟡 P1	Add unified `logs` command	Pending	Centralized log viewing across daemon, agents, swarms
🟢 P2	Add `upgrade` command	Pending	Self-update CLI to latest version
🟢 P2	Add interactive shell/REPL mode	Pending	`claude-flow shell` for interactive command execution

Implementation Plan

P0 - Critical (Next Release)

# Doctor command - diagnose system health
claude-flow doctor              # Full system check
claude-flow doctor --fix        # Auto-fix issues where possible
claude-flow doctor --component mcp  # Check specific component

# Shell completions
claude-flow completions bash > ~/.bash_completion.d/claude-flow
claude-flow completions zsh > ~/.zfunc/_claude-flow
claude-flow completions fish > ~/.config/fish/completions/claude-flow.fish

P1 - High Priority

# Unified logs command
claude-flow logs                # All logs
claude-flow logs --follow       # Tail logs
claude-flow logs --component daemon
claude-flow logs --level error

P2 - Nice to Have

# Self-update
claude-flow upgrade             # Upgrade to latest
claude-flow upgrade --check     # Check for updates
claude-flow upgrade --version 3.1.0

# Interactive shell
claude-flow shell               # Enter REPL
> swarm init mesh
> agent spawn coder
> memory search "patterns"

agentic-flow vs claude-flow Feature Comparison

Feature Matrix

Feature	agentic-flow	claude-flow	Integration Value
Core Agent System
Specialized Agents	66+	15 (hierarchical)	🟡
Agent Spawning	✅	✅	-
Agent Pool/Scaling	✅	✅	-
Orchestration
Swarm Coordination	✅	✅	-
Hive-Mind Consensus	❌	✅	-
Federation Hub	✅	❌	🔴 High
Ephemeral Agents	✅	❌	🔴 High
Transport
HTTP/SSE	✅	✅	-
QUIC (UDP)	✅	❌	🔴 High
WebSocket	✅	✅	-
AI Providers
Multi-Provider	✅	✅	-
Model Optimization	✅	❌	🔴 High
Provider Fallback	✅	❌	🔴 High
Cost Tracking	✅	❌	🟡 Medium
Proxy Server	✅	❌	🟡 Medium
Memory/Learning
ReasoningBank (WASM)	✅	❌	🔴 High
Embeddings CLI	✅	✅	-
HNSW Indexing	✅	✅	-
Memory Persistence	✅	✅	-
DevOps
Doctor Command	❌	✅	-
Shell Completions	❌	✅	-
Smart Errors	❌	✅	-
Background Workers	✅	✅	-
Worker Dispatch	✅	❌	🟡 Medium
Security
Security Scanning	❌	✅	-
CVE Detection	❌	✅	-
Claims/RBAC	❌	✅	-
Advanced
Neural Patterns	❌	✅	-
Self-Learning Hooks	✅	✅	-
Deployment Mgmt	❌	✅	-
Plugin System	❌	✅	-

High-Value Integration Opportunities

🔴 Tier 1: Highest Impact (Unique to agentic-flow)

1. QUIC Transport (50-70% faster)

# Integration target:
claude-flow transport quic --port 4433
claude-flow swarm start --transport quic  # 50-70% faster agent comms

2. Federation Hub (Ephemeral Agents)

# Integration target:
claude-flow federation start --port 9443
claude-flow federation spawn --tenant acme --lifetime 600

Value: Agents die but memories persist → learning across agent generations

3. Model Optimization (85% cost savings)

# Integration target:
claude-flow agent spawn -t coder --optimize --priority cost
claude-flow providers optimize --task "Build API" --budget 0.01

4. Provider Fallback (Enterprise resilience)

# Integration target:
claude-flow providers fallback configure --primary anthropic --fallback openrouter,onnx

5. ReasoningBank (WASM Learning Memory)

# Integration target:
claude-flow reasoningbank store "pattern" --reasoning "..."
claude-flow reasoningbank search "authentication patterns"

Value: 10-100x faster reasoning pattern storage vs JSON

🟡 Tier 2: Medium Impact

Feature	Description	Complexity
Proxy Server	Use any model with Claude Code via local proxy	🟡 Medium
Worker Dispatch	More granular worker control	🟢 Low
Embeddings Management	Better model management	🟢 Low

Integration Roadmap

Phase 1: Quick Wins (1 week) → 9.3 → 9.5

Feature	Effort	Impact	Source
Model Optimization flags	🟢 Low	🔴 High	New command
Provider fallback config	🟢 Low	🔴 High	New subcommand
Embeddings download/benchmark	🟢 Low	🟡 Med	Enhance existing

# After Phase 1:
claude-flow agent spawn -t coder --optimize --priority cost
claude-flow providers fallback configure --primary anthropic --fallback openrouter,onnx
claude-flow embeddings download all-MiniLM-L6-v2

Phase 2: Core Integration (2-3 weeks) → 9.5 → 9.7

Feature	Effort	Impact	Source
Federation Hub	🟡 Med	🔴 High	Port from agentic-flow
Ephemeral agents	🟡 Med	🔴 High	Part of federation
Proxy server	🟡 Med	🟡 Med	Port from agentic-flow
Worker dispatch triggers	🟢 Low	🟡 Med	Enhance daemon

# After Phase 2:
claude-flow federation start --port 9443
claude-flow federation spawn --lifetime 300 --task "Quick analysis"
claude-flow proxy start --provider openrouter
claude-flow daemon dispatch security-audit

Phase 3: Advanced (4-6 weeks) → 9.7 → 9.9

Feature	Effort	Impact	Source
QUIC Transport	🔴 High	🔴 High	Port QUIC module
ReasoningBank (WASM)	🔴 High	🔴 High	Port + WASM build
66 Agent configs	🟡 Med	🟡 Med	Copy + adapt

# After Phase 3:
claude-flow swarm start --transport quic  # 50-70% faster
claude-flow reasoningbank store "pattern" --reasoning "..."
claude-flow agent spawn -t security-analyst  # One of 66 types

Recommended Integration Approach

Option A: Dependency Approach (Fastest)

// claude-flow/package.json
{
  "dependencies": {
    "agentic-flow": "^2.0.3"
  }
}

Option B: Port Code (More Control)

# Copy specific modules:
- agentic-flow/src/transport/quic.ts → claude-flow/src/transport/
- agentic-flow/src/federation/ → claude-flow/src/federation/
- agentic-flow/src/reasoningbank/ → claude-flow/src/reasoningbank/

Option C: Unified Package (Long-term)

@claude-flow/core      # Shared primitives
@claude-flow/cli       # CLI (current)
@claude-flow/agents    # From agentic-flow's 66 agents
@claude-flow/transport # QUIC + HTTP + WebSocket

ruvector Integration Analysis

Package Overview

Package	Version	Description
`ruvector`	0.1.95	Main CLI + unified interface
`@ruvector/core`	0.1.30	Rust-native vector DB (52K+ inserts/sec)
`@ruvector/attention`	0.1.4	Flash Attention mechanisms
`@ruvector/sona`	0.1.5	Self-Optimizing Neural Architecture (LoRA, EWC++)
`@ruvector/gnn`	0.1.22	Graph Neural Networks

Feature Overlap Analysis

claude-flow ALREADY HAS (via @claude-flow/embeddings):

Feature	claude-flow	ruvector	Status
ONNX Embeddings	✅ agentic-flow (~3ms)	✅ @ruvector/core	Equivalent
Local Embeddings	✅ all-MiniLM-L6-v2	✅ all-MiniLM-L6-v2	Equivalent
HNSW Indexing	✅ @claude-flow/memory	✅ @ruvector/core	Equivalent
Persistent Cache	✅ SQLite + LRU	✅ Memory cache	Equivalent
Hyperbolic Embeddings	✅ Poincaré ball	❌	claude-flow ahead
Document Chunking	✅ 4 strategies	❌	claude-flow ahead
Normalization	✅ L2, L1, min-max, z-score	❌	claude-flow ahead
Neural Substrate	✅ Drift, memory physics	❌	claude-flow ahead

Unique ruvector Features (Integration Candidates)

🔴 Tier 1: High Value (claude-flow lacks these)

Feature	ruvector Source	Integration Target	Value
Q-Learning Agent Router	`hooks_route`	`hooks route --task`	80%+ accuracy, learns patterns
AST Analysis	`hooks_ast_analyze`	`analyze --ast`	Symbol extraction, complexity
Diff Classification	`hooks_diff_analyze`	`analyze --diff --risk`	Change risk scoring
Coverage Routing	`hooks_coverage_route`	`hooks route --coverage-aware`	Test-aware agent selection

🟡 Tier 2: Medium Value

Feature	ruvector Source	Integration Target
Co-edit Prediction	Git history analysis	`predict --coedits`
Security Patterns	`hooks_security_scan`	`security scan --patterns`

🟢 Tier 3: Nice to Have

Feature	ruvector Source	Use Case
MinCut Boundaries	`hooks_graph_mincut`	Code organization
Louvain Communities	`hooks_graph_cluster`	Module detection
GNN Layers	`@ruvector/gnn`	Graph analysis

Note: Flash Attention, SONA Learning, HNSW, and ONNX embeddings are already in claude-flow via agentic-flow.

MCP Tools from ruvector (Unique Only)

# Add ruvector MCP server (for unique features)
claude mcp add ruvector-mcp -- npx ruvector mcp-server

Unique Tools Worth Integrating:

hooks_route, hooks_route_enhanced — Q-Learning agent routing (80%+ accuracy) ✅
hooks_ast_analyze, hooks_ast_complexity — Code structure analysis ✅
hooks_diff_analyze, hooks_diff_classify — Change classification ✅
hooks_coverage_route, hooks_coverage_suggest — Test-aware routing ✅
hooks_graph_mincut, hooks_graph_cluster — Code boundaries ✅

Already in claude-flow (skip):

hooks_rag_context — Use @claude-flow/memory instead
hooks_attention_info — Use @claude-flow/neural instead
Embeddings tools — Use @claude-flow/embeddings instead

Integration Approach

Recommended: Option A - Add as Optional Dependency (for unique features only)

// @claude-flow/cli/package.json
{
  "optionalDependencies": {
    "ruvector": "^0.1.95"
  }
}

CLI Wrappers (unique ruvector features):

# Q-Learning agent routing (unique to ruvector)
claude-flow route "task" --q-learning          # Uses hooks_route

# AST analysis (unique to ruvector)
claude-flow analyze ast src/                   # Uses hooks_ast_analyze

# Diff classification (unique to ruvector)
claude-flow analyze diff --risk                # Uses hooks_diff_analyze

# Coverage-aware routing (unique to ruvector)
claude-flow route "task" --coverage-aware      # Uses hooks_coverage_route

Already in claude-flow (DO NOT import from ruvector):

claude-flow embeddings generate --local        # Uses @claude-flow/embeddings (ONNX)
claude-flow memory search --semantic "query"   # Uses @claude-flow/memory (HNSW)

ruvector Integration Roadmap

Phase 1: Q-Learning Router (1-2 days)

Add ruvector as optional dependency
Implement hooks route --q-learning wrapper
Add info --ruvector command for capability detection

Phase 2: Code Intelligence (1 week)

Integrate AST analysis commands (analyze ast)
Implement diff classification (analyze diff --risk)
Add coverage-aware routing (route --coverage-aware)

Phase 3: Graph Analysis (2 weeks)

Implement MinCut boundaries (analyze --boundaries)
Add Louvain community detection (analyze --modules)
Integrate GNN layers for dependency graphs

Note: SONA, Flash Attention, HNSW already in claude-flow - no need to import.