🌊 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 edits
rUvde10bd130d74
diff --git a/README.md b/README.mdindex 6307d7e2c3..285ae9eb19 100644--- a/README.md+++ b/README.md@@ -220,15 +220,18 @@ Claude-Flow v3 introduces **self-learning neural capabilities** that no other ag| **Swarm Topologies** | ✅ 4 (mesh, hierarchical, etc.) | 1 | 1 | 1 | 1 |**Key differentiators:**-- **SONA (Self-Optimizing Neural Architecture)** - Learns optimal agent routing from trajectory outcomes with <0.05ms adaptation-- **EWC++ (Elastic Weight Consolidation)** - Prevents catastrophic forgetting when learning new patterns-- **MoE (Mixture of Experts)** - 8 specialized experts with dynamic gating network for intelligent task routing-- **Flash Attention** - 2.49x-7.47x speedup for attention computations-- **Hyperbolic Embeddings** - Poincaré ball model for hierarchical data representation (better for code structure)-- **LoRA (Low-Rank Adaptation)** - 128x memory compression for efficient fine-tuning-- **Int8 Quantization** - 3.92x memory reduction with minimal accuracy loss-- **Claims System** - Human-agent work ownership with handoff protocols-- **Byzantine Consensus** - Fault-tolerant coordination even with malicious agents++| | Feature | What It Does | Technical Details |+|---|---------|--------------|-------------------|+| 🧠 | **SONA** | System gets smarter over time by learning which agents work best for each task | Self-Optimizing Neural Architecture with <0.05ms adaptation from trajectory outcomes |+| 🔒 | **EWC++** | Remembers successful patterns without forgetting old ones when learning new things | Elastic Weight Consolidation prevents catastrophic forgetting during continuous learning |+| 🎯 | **MoE** | Automatically picks the right specialist for each job | Mixture of 8 Experts with dynamic gating network for intelligent task routing |+| ⚡ | **Flash Attention** | Makes thinking 2-7x faster | 2.49x-7.47x speedup for attention computations using optimized CUDA kernels |+| 🌐 | **Hyperbolic Embeddings** | Better understands code hierarchies (folders, classes, functions) | Poincaré ball model captures hierarchical relationships in lower dimensions |+| 📦 | **LoRA** | Runs on less powerful machines by compressing the AI brain | Low-Rank Adaptation achieves 128x memory compression for efficient fine-tuning |+| 🗜️ | **Int8 Quantization** | Uses 4x less memory with almost no quality loss | 3.92x memory reduction via 8-bit integer quantization with calibration |+| 🤝 | **Claims System** | Tracks who's working on what, enables smooth handoffs | Human-agent work ownership with claim/release/handoff protocols |+| 🛡️ | **Byzantine Consensus** | Keeps working even if some agents misbehave | Fault-tolerant coordination tolerating up to f < n/3 malicious agents |### Intelligent 3-Tier Model Routing (ADR-026)