Skip to content

armgit / ruflo

Public Imported from GitHub
0

🌊 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

Checkpoint: File edits

rUvf375a70cbd9f
diff --git a/README.md b/README.md
index 371aaf1491..96bd2479d7 100644
--- a/README.md
+++ b/README.md
@@ -19,6 +19,118 @@
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.
+### Architecture Flow
+
+```mermaid
+flowchart TB
+ subgraph User["User Layer"]
+ CC[Claude Code]
+ CLI[CLI Commands]
+ end
+
+ subgraph Orchestration["Orchestration Layer"]
+ MCP[MCP Server<br/>27+ Tools]
+ Router[Intelligent Router<br/>3-Tier Model Selection]
+ Hooks[Self-Learning Hooks<br/>27 Lifecycle Events]
+ end
+
+ subgraph Coordination["Coordination Layer"]
+ Queen[Queen Coordinator]
+ Swarm[Swarm Topologies<br/>Mesh • Hierarchical • Ring • Star]
+ Consensus[Consensus Protocols<br/>Raft • Byzantine • Gossip • CRDT]
+ end
+
+ subgraph Agents["Agent Layer (54+)"]
+ Core[Core Agents<br/>coder • reviewer • tester]
+ Specialized[Specialized Agents<br/>security • performance • ml]
+ GitHub[GitHub Agents<br/>pr-manager • issue-tracker]
+ end
+
+ subgraph Intelligence["Intelligence Layer"]
+ SONA[SONA Neural<br/>Self-Optimizing Architecture]
+ MoE[Mixture of Experts<br/>8 Specialized Experts]
+ ReasoningBank[ReasoningBank<br/>Pattern Learning]
+ end
+
+ subgraph Memory["Memory Layer"]
+ HNSW[HNSW Index<br/>150x-12,500x Faster]
+ AgentDB[AgentDB<br/>Vector Storage]
+ Cache[LRU Cache<br/>Cross-Session]
+ end
+
+ subgraph Providers["Provider Layer"]
+ Anthropic[Anthropic<br/>Claude 3.5/4]
+ OpenAI[OpenAI<br/>GPT-4o/o1]
+ Google[Google<br/>Gemini 2.0]
+ Local[Local<br/>Ollama/Llama]
+ end
+
+ CC --> MCP
+ CLI --> MCP
+ MCP --> Router
+ Router --> Hooks
+ Hooks --> Queen
+ Queen --> Swarm
+ Swarm --> Consensus
+ Consensus --> Core
+ Consensus --> Specialized
+ Consensus --> GitHub
+ Core --> SONA
+ Specialized --> MoE
+ GitHub --> ReasoningBank
+ SONA --> HNSW
+ MoE --> AgentDB
+ ReasoningBank --> Cache
+ HNSW --> Anthropic
+ AgentDB --> OpenAI
+ Cache --> Google
+ Cache --> Local
+
+ style User fill:#e1f5fe
+ style Orchestration fill:#f3e5f5
+ style Coordination fill:#fff3e0
+ style Agents fill:#e8f5e9
+ style Intelligence fill:#fce4ec
+ style Memory fill:#e0f2f1
+ style Providers fill:#fff8e1
+```
+
+### Data Flow
+
+```mermaid
+sequenceDiagram
+ participant U as User/Claude Code
+ participant R as Router (3-Tier)
+ participant S as Swarm
+ participant A as Agents
+ participant M as Memory (HNSW)
+ participant L as LLM Provider
+
+ U->>R: Task Request
+ R->>R: Complexity Analysis
+ alt Simple Transform
+ R->>U: Agent Booster (WASM) <1ms
+ else Medium Complexity
+ R->>S: Route to Haiku/Sonnet
+ S->>A: Spawn Agents
+ A->>M: Query Patterns
+ M->>A: Similar Patterns (150x faster)
+ A->>L: LLM Request
+ L->>A: Response
+ A->>M: Store New Pattern
+ A->>S: Result
+ S->>U: Synthesized Output
+ else High Complexity
+ R->>S: Route to Opus
+ S->>A: Spawn Specialized Swarm
+ A->>A: Byzantine Consensus
+ A->>M: Deep Pattern Search
+ A->>L: Multi-turn Reasoning
+ A->>S: Coordinated Result
+ S->>U: Final Output
+ end
+```
+
### Claude Code: With vs Without Claude-Flow
| Capability | Claude Code Alone | Claude Code + Claude-Flow |