🌊 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
rUvd9af27569874
diff --git a/CLAUDE.md b/CLAUDE.mdindex 2dc808534c..deb87aae2f 100644--- a/CLAUDE.md+++ b/CLAUDE.md@@ -35,24 +35,18 @@----## Automatic Swarm Orchestration+## Swarm Orchestration-**When starting work on complex tasks, Claude Code MUST automatically:**--1. **Initialize the swarm** using MCP tools-2. **Spawn concurrent agents** using Claude Code's Task tool-3. **Coordinate via hooks** and memory--### CRITICAL: MCP + Task Tool in SAME Message+- MUST initialize the swarm using MCP tools when starting complex tasks+- MUST spawn concurrent agents using Claude Code's Task tool+- Never use MCP tools alone for execution — Task tool agents do the actual work-**When user says "spawn swarm" or requests complex work, Claude Code MUST in ONE message:**-1. Call MCP tools to initialize coordination-2. **IMMEDIATELY** call Task tool to spawn REAL working agents-3. Both MCP and Task calls must be in the SAME response+### MCP + Task Tool in SAME Message-- Never use MCP tools alone for execution — Task tool agents do the actual work+- MUST call MCP tools AND Task tool in ONE message for complex work+- Always call MCP first, then IMMEDIATELY call Task tool to spawn agents-### Intelligent 3-Tier Model Routing (ADR-026)+### 3-Tier Model Routing (ADR-026)| Tier | Handler | Latency | Cost | Use Cases ||------|---------|---------|------|-----------|@@ -60,55 +54,27 @@| **2** | Haiku | ~500ms | $0.0002 | Simple tasks, low complexity (<30%) || **3** | Sonnet/Opus | 2-5s | $0.003-0.015 | Complex reasoning, architecture, security (>30%) |-**When you see these recommendations:**--1. `[AGENT_BOOSTER_AVAILABLE]` → Use the Edit tool directly for simple code transforms- - Intent types: `var-to-const`, `add-types`, `add-error-handling`, `async-await`, `add-logging`, `remove-console`--2. `[TASK_MODEL_RECOMMENDATION] Use model="X"` → Use that model in Task tool:-```javascript-Task({- prompt: "...",- subagent_type: "coder",- model: "opus" // ← USE THE RECOMMENDED MODEL-})-```--**Model Selection Logic:**-| Complexity | Model | Use For |-|------------|-------|---------|-| Agent Booster intent detected | **Skip LLM** | var→const, add-types, remove-console (352x faster) |-| High (architecture, system design, security) | **opus** | Complex reasoning, multi-step planning |-| Medium (features, refactoring, debugging) | **sonnet** | Balanced capability and speed |-| Low (formatting, simple fixes, docs) | **haiku** | Fast, cost-effective tasks |-- Always check for `[AGENT_BOOSTER_AVAILABLE]` or `[TASK_MODEL_RECOMMENDATION]` before spawning agents+- Use Edit tool directly when `[AGENT_BOOSTER_AVAILABLE]` — intent types: `var-to-const`, `add-types`, `add-error-handling`, `async-await`, `add-logging`, `remove-console`### Anti-Drift Coding Swarm (PREFERRED DEFAULT)-- ALWAYS use this configuration for coding swarms:+- ALWAYS use hierarchical topology for coding swarms+- Keep maxAgents at 6-8 for tight coordination+- Use specialized strategy for clear role boundaries+- Use `raft` consensus for hive-mind (leader maintains authoritative state)+- Run frequent checkpoints via `post-task` hooks+- Keep shared memory namespace for all agents+- Keep task cycles short with verification gates```javascriptmcp__ruv-swarm__swarm_init({- topology: "hierarchical", // Single coordinator enforces alignment- maxAgents: 8, // Smaller team = less drift surface- strategy: "specialized" // Clear roles reduce ambiguity+ topology: "hierarchical",+ maxAgents: 8,+ strategy: "specialized"})```-| Choice | Anti-Drift Benefit |-|--------|-------------------|-| **hierarchical** | Coordinator validates each output against goal, catches divergence early |-| **maxAgents: 6-8** | Fewer agents = less coordination overhead, easier alignment |-| **specialized** | Clear boundaries — each agent knows exactly what to do, no overlap |--- Use `raft` consensus for hive-mind (leader maintains authoritative state)--**Additional anti-drift measures:**-- Run frequent checkpoints via `post-task` hooks-- Keep shared memory namespace for all agents-- Keep task cycles short with verification gates-### Auto-Start Swarm ProtocolWhen the user requests a complex task (multi-file changes, feature implementation, refactoring), **immediately execute this pattern in a SINGLE message:**