🌊 The original agent harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversational AI systems. Features adaptive memory, self-learning intelligence, federation, vector RAG integration, and native Claude Code / Codex / Hermes and many more Integrated
TypeScript agentic-aiagentic-frameworkagentic-workflowagentsai-agentsai-assistantai-skillsautonomous-agentsclaudecodexdsh-pluginharnessmcp-servermulti-agentmulti-agent-systemsnpmskillsswarmswarm-intelligencetypescript
checkpoint: File edit:
rUve402bcdd438c
diff --git a/README.md b/README.mdindex 3b7eddb3e3..cbce99c54b 100644--- a/README.md+++ b/README.md@@ -19,16 +19,20 @@Claude-Flow is a comprehensive AI agent orchestration framework that transforms Claude Code into a powerful multi-agent development platform. It enables teams to deploy, coordinate, and optimize specialized AI agents working together on complex software engineering tasks.-### What Makes Claude-Flow Different--| Capability | Traditional Agents | Claude-Flow v3 |-|------------|-------------------|----------------|-| **Agent Coordination** | Single agent, sequential | 15 concurrent agents, 6 swarm topologies |-| **Learning** | Static prompts | Self-learning with HNSW (150x faster retrieval) |-| **Fault Tolerance** | None | Byzantine, Raft, Gossip, CRDT consensus |-| **Memory** | Session-only | Persistent AgentDB with vector search |-| **Performance** | Baseline | 2.8-4.4x speed improvement |-| **Security** | Basic | CVE-hardened, input validation, sandboxing |+### Claude Code: With vs Without Claude-Flow++| Capability | Claude Code Alone | Claude Code + Claude-Flow |+|------------|-------------------|---------------------------|+| **Agents** | 1 general-purpose agent | 54+ specialized agents (coder, tester, reviewer, security, DevOps...) |+| **Parallelism** | Sequential tasks only | 15 agents working simultaneously |+| **Memory** | Forgets between sessions | Remembers patterns, learns what works, improves over time |+| **Complex Tasks** | Manual breakdown required | Automatic task decomposition across agent teams |+| **Code Review** | Single-pass review | Multi-agent review (security + performance + style + tests) |+| **Testing** | Write tests when asked | Automatic test generation with TDD patterns |+| **LLM Provider** | Anthropic only | Claude, GPT-4, Gemini, Cohere, Ollama (with cost optimization) |+| **Failure Handling** | Restart from scratch | Automatic recovery, task retry, agent failover |+| **Security** | Standard protections | CVE-hardened, sandboxed commands, validated inputs |+| **Speed** | Baseline | 2.8-4.4x faster on complex tasks |### Key Capabilities