🌊 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: Edit /workspaces/claude-code-flow/src/cli/simple-commands/swarm.js
rUv567922e159e8
diff --git a/src/cli/simple-commands/swarm.js b/src/cli/simple-commands/swarm.jsindex 8ed18a82c3..8c4e56d857 100644--- a/src/cli/simple-commands/swarm.js+++ b/src/cli/simple-commands/swarm.js@@ -430,36 +430,48 @@ If you need to do X operations, they should be in 1 message, not X messages.🎯 MANDATORY PATTERNS FOR CLAUDE-FLOW SWARMS:-1️⃣ **SWARM INITIALIZATION** - Everything in ONE BatchTool:+1️⃣ **SWARM INITIALIZATION** - Use Claude Code's Task Tool for Agents:++Step A: Optional MCP Coordination Setup (Single Message):\`\`\`javascript-[Single Message with Multiple Tools]:- // Spawn ALL agents at once+[MCP Tools - Coordination ONLY]:+ // Set up coordination topology (OPTIONAL)+ mcp__claude-flow__swarm_init {"topology": "mesh", "maxAgents": ${maxAgents}}mcp__claude-flow__agent_spawn {"type": "coordinator", "name": "SwarmLead"}- mcp__claude-flow__agent_spawn {"type": "researcher", "name": "DataAnalyst"}- mcp__claude-flow__agent_spawn {"type": "coder", "name": "BackendDev"}- mcp__claude-flow__agent_spawn {"type": "coder", "name": "FrontendDev"}- mcp__claude-flow__agent_spawn {"type": "tester", "name": "QAEngineer"}-- // Initialize ALL memory keysmcp__claude-flow__memory_store {"key": "swarm/objective", "value": "${objective}"}- mcp__claude-flow__memory_store {"key": "swarm/config", "value": {"strategy": "${strategy}", "mode": "${mode}"}}-- // Create task hierarchy- mcp__claude-flow__task_create {"name": "${objective}", "type": "parent", "id": "main"}- mcp__claude-flow__task_create {"name": "Research Phase", "parent": "main"}- mcp__claude-flow__task_create {"name": "Design Phase", "parent": "main"}- mcp__claude-flow__task_create {"name": "Implementation", "parent": "main"}+ mcp__claude-flow__memory_store {"key": "swarm/config", "value": {"strategy": "${strategy}"}}+\`\`\`++Step B: REQUIRED - Claude Code Task Tool for ACTUAL Agent Execution (Single Message):+\`\`\`javascript+[Claude Code Task Tool - CONCURRENT Agent Spawning]:+ // Spawn ALL agents using Task tool in ONE message+ Task("Coordinator", "Lead swarm coordination. Use hooks for memory sharing.", "coordinator")+ Task("Researcher", "Analyze requirements and patterns. Coordinate via hooks.", "researcher")+ Task("Backend Dev", "Implement server-side features. Share progress via hooks.", "coder")+ Task("Frontend Dev", "Build UI components. Sync with backend via memory.", "coder")+ Task("QA Engineer", "Create and run tests. Report findings via hooks.", "tester")- // Initialize comprehensive todo list+ // Batch ALL todos in ONE TodoWrite call (5-10+ todos)TodoWrite {"todos": [{"id": "1", "content": "Initialize ${maxAgents} agent swarm", "status": "completed", "priority": "high"},{"id": "2", "content": "Analyze: ${objective}", "status": "in_progress", "priority": "high"},{"id": "3", "content": "Design architecture", "status": "pending", "priority": "high"},- {"id": "4", "content": "Implement solution", "status": "pending", "priority": "high"},- {"id": "5", "content": "Test and validate", "status": "pending", "priority": "medium"}+ {"id": "4", "content": "Implement backend", "status": "pending", "priority": "high"},+ {"id": "5", "content": "Implement frontend", "status": "pending", "priority": "high"},+ {"id": "6", "content": "Write unit tests", "status": "pending", "priority": "medium"},+ {"id": "7", "content": "Integration testing", "status": "pending", "priority": "medium"},+ {"id": "8", "content": "Performance optimization", "status": "pending", "priority": "low"},+ {"id": "9", "content": "Documentation", "status": "pending", "priority": "low"}]}\`\`\`+⚠️ CRITICAL: Claude Code's Task tool does the ACTUAL work!+- MCP tools = Coordination setup only+- Task tool = Spawns agents that execute real work+- ALL agents MUST be spawned in ONE message+- ALL todos MUST be batched in ONE TodoWrite call+2️⃣ **TASK COORDINATION** - Batch ALL assignments:\`\`\`javascript[Single Message]:diff --git a/test-claude-md-update/.claude/checkpoints/1755801153.json b/test-claude-md-update/.claude/checkpoints/1755801153.jsonnew file mode 100644index 0000000000..7b9054be0c--- /dev/null+++ b/test-claude-md-update/.claude/checkpoints/1755801153.json@@ -0,0 +1 @@+{"branch": "checkpoint/pre-edit-20250821-183233", "file": "/workspaces/claude-code-flow/src/cli/simple-commands/hive-mind.js", "timestamp": "2025-08-21T18:32:33Z", "type": "pre-edit", "original_branch": "alpha91"}diff --git a/test-claude-md-update/.claude/checkpoints/1755801155.json b/test-claude-md-update/.claude/checkpoints/1755801155.jsonnew file mode 100644index 0000000000..b009ed86ae--- /dev/null+++ b/test-claude-md-update/.claude/checkpoints/1755801155.json@@ -0,0 +1 @@+{"tag": "checkpoint-20250821-183234", "file": "/workspaces/claude-code-flow/src/cli/simple-commands/hive-mind.js", "timestamp": "2025-08-21T18:32:35Z", "type": "post-edit", "branch": "alpha91", "diff_summary": " .claude-flow/metrics/system-metrics.json | 96 ++++++++++++++++++++++ claude-flow-wiki | 0 src/cli/simple-commands/hive-mind.js | 28 +++++-- .../.claude/checkpoints/1755801135.json | 1 + .../.claude/checkpoints/1755801137.json | 1 + 5 files changed, 121 insertions(+), 5 deletions(-) "}