🌟 Overview
Claude-Flow v2.7 is an enterprise-grade AI orchestration platform that combines hive-mind swarm intelligence , persistent memory , and 100+ advanced MCP tools to revolutionize AI-powered development workflows.
🎯 Key Features
🎨 25 Claude Skills : Natural language-activated skills for development, GitHub, memory, and automation🚀 AgentDB v1.3.9 Integration : 96x-164x faster vector search with semantic understanding (PR #830)🧠 Hybrid Memory System : AgentDB + ReasoningBank with automatic fallback🔍 Semantic Vector Search : HNSW indexing (O(log n)) + 9 RL algorithms🐝 Hive-Mind Intelligence : Queen-led AI coordination with specialized worker agents🔧 100 MCP Tools : Comprehensive toolkit for swarm orchestration and automation🔄 Dynamic Agent Architecture (DAA) : Self-organizing agents with fault tolerance💾 Persistent Memory : 150x faster search, 4-32x memory reduction (quantization)🪝 Advanced Hooks System : Automated workflows with pre/post operation hooks📊 GitHub Integration : 6 specialized modes for repository management🌐 Flow Nexus Cloud : E2B sandboxes, AI swarms, challenges, and marketplace
🔥 Revolutionary AI Coordination : Build faster, smarter, and more efficiently with AI-powered development orchestration
🆕 NEW: AgentDB Integration : 96x-164x performance boost with semantic vector search, reflexion memory, and skill library auto-consolidation
⚡ Quick Start
📋 Prerequisites
Node.js 18+ (LTS recommended)npm 9+ or equivalent package managerWindows users : See Windows Installation Guide for special instructions
⚠️ IMPORTANT : Claude Code must be installed first:
# 1. Install Claude Code globally npm install -g @anthropic-ai/claude-code # 2. (Optional) Skip permissions check for faster setup claude --dangerously-skip-permissions 🚀 Install Latest Alpha
# NPX (recommended - always latest) npx claude-flow@alpha init --force npx claude-flow@alpha --help # Or install globally npm install -g claude-flow@alpha claude-flow --version # v2.7.0-alpha.10 🎨 Skills System
Claude-Flow includes 25 specialized skills that activate automatically via natural language - no commands to memorize:
# Just describe what you want - skills activate automatically "Let's pair program on this feature" → pair-programming skill "Review this PR for security issues" → github-code-review skill "Use vector search to find similar code" → agentdb-vector-search skill "Create a swarm to build this API" → swarm-orchestration skill Skill Categories:
Development & Methodology (3) - SPARC, pair programming, skill builderIntelligence & Memory (6) - AgentDB integration with 150x-12,500x performanceSwarm Coordination (3) - Multi-agent orchestration and hive-mindGitHub Integration (5) - PR review, workflows, releases, multi-repoAutomation & Quality (4) - Hooks, verification, performance analysisFlow Nexus Platform (3) - Cloud sandboxes and neural training
📚 Complete Skills Tutorial
🆕 What's New in v2.7.0-alpha.10
✅ Semantic Search Fixed
Critical bug fix for semantic search returning 0 results:
✅ Fixed stale compiled code (dist-cjs/ now uses Node.js backend)
✅ Fixed result mapping for retrieveMemories() flat structure
✅ Fixed parameter mismatch (namespace vs domain)
✅ 2-3ms query latency with hash embeddings
✅ Works without API keys (deterministic 1024-dim embeddings)
Node.js Backend : Replaced WASM with SQLite + better-sqlite3Persistent Storage : All memories saved to .swarm/memory.dbSemantic Search : MMR ranking with 4-factor scoringDatabase Tables : patterns, embeddings, trajectories, linksPerformance : 2ms queries, 400KB per pattern with embeddings
# Semantic search now fully functional npx claude-flow@alpha memory store test "API configuration" --namespace semantic --reasoningbank npx claude-flow@alpha memory query "configuration" --namespace semantic --reasoningbank # ✅ Found 3 results (semantic search) in 2ms 📚 Release Notes : v2.7.0-alpha.10
🧠 Memory System Commands
Revolutionary Performance Improvements:
Vector Search : 96x faster (9.6ms → <0.1ms)Batch Operations : 125x fasterLarge Queries : 164x fasterMemory Usage : 4-32x reduction via quantization
# Semantic vector search (understands meaning, not just keywords) npx claude-flow@alpha memory vector-search "user authentication flow" \ --k 10 --threshold 0.7 --namespace backend # Store with vector embedding for semantic search npx claude-flow@alpha memory store-vector api_design "REST endpoints" \ --namespace backend --metadata '{"version":"v2"}' # Get AgentDB integration status and capabilities npx claude-flow@alpha memory agentdb-info # Installation (hybrid mode - 100% backward compatible) npm install [email protected] New Features:
✅ Semantic vector search (HNSW indexing, O(log n))
✅ 9 RL algorithms (Q-Learning, PPO, MCTS, Decision Transformer)
✅ Reflexion memory (learn from past experiences)
✅ Skill library (auto-consolidate successful patterns)
✅ Causal reasoning (understand cause-effect relationships)
✅ Quantization (binary 32x, scalar 4x, product 8-16x reduction)
✅ 100% backward compatible (hybrid mode with graceful fallback)
Documentation : docs/agentdb/PRODUCTION_READINESS.md | PR : #830
ReasoningBank (Legacy SQLite Memory - Still Supported) # Store memories with pattern matching npx claude-flow@alpha memory store api_key "REST API configuration" \ --namespace backend --reasoningbank # Query with pattern search (2-3ms latency) npx claude-flow@alpha memory query "API config" \ --namespace backend --reasoningbank # ✅ Found 3 results (pattern matching) # List all memories npx claude-flow@alpha memory list --namespace backend --reasoningbank # Check status and statistics npx claude-flow@alpha memory status --reasoningbank # ✅ Total memories: 30 # Embeddings: 30 # Storage: .swarm/memory.db Features:
✅ No API Keys Required : Hash-based embeddings (1024 dimensions)
✅ Persistent Storage : SQLite database survives restarts
✅ Pattern Matching : LIKE-based search with similarity scoring
✅ Namespace Isolation : Organize memories by domain
✅ Fast Queries : 2-3ms average latency
✅ Process Cleanup : Automatic database closing
Optional Enhanced Embeddings:
# For better semantic accuracy with text-embedding-3-small (1536 dimensions) # Set OPENAI environment variable (see ReasoningBank documentation) 🐝 Swarm Orchestration
Quick Swarm Commands # Quick task execution (recommended) npx claude-flow@alpha swarm "build REST API with authentication" --claude # Multi-agent coordination npx claude-flow@alpha swarm init --topology mesh --max-agents 5 npx claude-flow@alpha swarm spawn researcher "analyze API patterns" npx claude-flow@alpha swarm spawn coder "implement endpoints" npx claude-flow@alpha swarm status Hive-Mind for Complex Projects # Initialize hive-mind system npx claude-flow@alpha hive-mind wizard npx claude-flow@alpha hive-mind spawn "build enterprise system" --claude # Session management npx claude-flow@alpha hive-mind status npx claude-flow@alpha hive-mind resume session-xxxxx When to Use:
Feature swarmhive-mindBest For Quick tasks Complex projects Setup Instant Interactive wizard Memory Task-scoped Project-wide SQLite Sessions Temporary Persistent + resume
Setup MCP Servers # Add Claude Flow MCP server (required) claude mcp add claude-flow npx claude-flow@alpha mcp start # Optional: Enhanced coordination claude mcp add ruv-swarm npx ruv-swarm mcp start # Optional: Cloud features (requires registration) claude mcp add flow-nexus npx flow-nexus@latest mcp start Core Tools:
swarm_init, agent_spawn, task_orchestratememory_usage, memory_searchneural_status, neural_train, neural_patterns
Memory Tools:
mcp__claude-flow__memory_usage - Store/retrieve persistent memorymcp__claude-flow__memory_search - Pattern-based search
GitHub Tools:
github_repo_analyze, github_pr_manage, github_issue_track
Performance Tools:
benchmark_run, performance_report, bottleneck_analyze
📚 Full Reference : MCP Tools Documentation
🪝 Advanced Hooks System
Automated Workflow Enhancement Claude-Flow automatically configures hooks for enhanced operations:
# Auto-configures hooks during init npx claude-flow@alpha init --force Available Hooks Pre-Operation:
pre-task: Auto-assigns agents by complexitypre-edit: Validates files and prepares resourcespre-command: Security validation
Post-Operation:
post-edit: Auto-formats codepost-task: Trains neural patternspost-command: Updates memory
Session Management:
session-start: Restores previous contextsession-end: Generates summariessession-restore: Loads memory
🎯 Common Workflows
Pattern 1: Single Feature Development # Initialize once per feature npx claude-flow@alpha init --force npx claude-flow@alpha hive-mind spawn "Implement authentication" --claude # Continue same feature (reuse hive) npx claude-flow@alpha memory query "auth" --recent npx claude-flow@alpha swarm "Add password reset" --continue-session Pattern 2: Multi-Feature Project # Project initialization npx claude-flow@alpha init --force --project-name "my-app" # Feature 1: Authentication npx claude-flow@alpha hive-mind spawn "auth-system" --namespace auth --claude # Feature 2: User management npx claude-flow@alpha hive-mind spawn "user-mgmt" --namespace users --claude Pattern 3: Research & Analysis # Start research session npx claude-flow@alpha hive-mind spawn "Research microservices" \ --agents researcher,analyst --claude # Check learned knowledge npx claude-flow@alpha memory stats npx claude-flow@alpha memory query "microservices patterns" --reasoningbank
84.8% SWE-Bench solve rate - Industry-leading problem-solving32.3% token reduction - Efficient context management2.8-4.4x speed improvement - Parallel coordination96x-164x faster search - 🆕 AgentDB vector search (9.6ms → <0.1ms)4-32x memory reduction - 🆕 AgentDB quantization2-3ms query latency - ReasoningBank pattern search (legacy)64 specialized agents - Complete development ecosystem100 MCP tools - Comprehensive automation toolkit180 AgentDB tests - >90% coverage, production-ready
📚 Documentation
📖 Core Documentation 🚀 Release Notes & Changelogs 🛠️ Advanced Topics ⚙️ Configuration & Setup 🚀 Roadmap & Targets
✅ Semantic search fix (v2.7.0-alpha.10)
✅ ReasoningBank Node.js backend
✅ AgentDB v1.3.9 integration (PR #830) - 96x-164x performance boost
🔄 AgentDB production deployment (Q4 2025)
🔄 Enhanced embedding models
🔄 Multi-user collaboration features
Q1 2026
Advanced neural pattern recognition
Cloud swarm coordination
Real-time agent communication
Enterprise SSO integration
Growth Targets
5K+ GitHub stars, 50K npm downloads/month
$25K MRR, 15 enterprise customers
90%+ error prevention
30+ minutes saved per developer per week
Star History
📄 License
MIT License - see LICENSE for details
Built with ❤️ by rUv | Powered by Revolutionary AI
v2.7.0-alpha.10 - Semantic Search Fixed + ReasoningBank Node.js Backend
