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: Edit /workspaces/claude-code-flow/.claude/commands/agents/agent-spawning.md

rUv3213fde395b7
diff --git a/.claude/checkpoints/1758908507.json b/.claude/checkpoints/1758908507.json
new file mode 100644
index 0000000000..9cfbe0850d
--- /dev/null
+++ b/.claude/checkpoints/1758908507.json
@@ -0,0 +1 @@
+{"branch": "checkpoint/pre-edit-20250926-174146", "file": "/workspaces/claude-code-flow/.claude/commands/hive-mind/hive-mind-spawn.md", "timestamp": "2025-09-26T17:41:47Z", "type": "pre-edit", "original_branch": "alpha-125"}
diff --git a/.claude/checkpoints/1758908508.json b/.claude/checkpoints/1758908508.json
new file mode 100644
index 0000000000..5bca5ca8a9
--- /dev/null
+++ b/.claude/checkpoints/1758908508.json
@@ -0,0 +1 @@
+{"tag": "checkpoint-20250926-174148", "file": "/workspaces/claude-code-flow/.claude/commands/hive-mind/hive-mind-spawn.md", "timestamp": "2025-09-26T17:41:48Z", "type": "post-edit", "branch": "alpha-125", "diff_summary": " .claude-flow/metrics/performance.json | 2 +- .claude-flow/metrics/system-metrics.json | 48 +++++++++++++++ .claude-flow/metrics/task-metrics.json | 6 +- .claude/checkpoints/task-1758908385.json | 1 + .claude/commands/hive-mind/hive-mind-spawn.md | 88 +++++++++++++++++++++++++-- claude-flow-wiki | 0 6 files changed, 137 insertions(+), 8 deletions(-) "}
diff --git a/.claude/commands/agents/agent-spawning.md b/.claude/commands/agents/agent-spawning.md
index 38c8581d4c..e47790dee7 100644
--- a/.claude/commands/agents/agent-spawning.md
+++ b/.claude/commands/agents/agent-spawning.md
@@ -1,28 +1,103 @@
# agent-spawning
-Guide to spawning agents with Claude Code's Task tool.
+Guide to spawning agents with Claude Code's Task tool and MCP coordination.
## Using Claude Code's Task Tool
**CRITICAL**: Always use Claude Code's Task tool for actual agent execution:
```javascript
-// Spawn ALL agents in ONE message
-Task("Researcher", "Analyze requirements...", "researcher")
-Task("Coder", "Implement features...", "coder")
-Task("Tester", "Create tests...", "tester")
+// Spawn ALL agents in ONE message for concurrent execution
+Task("Researcher", "Analyze requirements and write findings to memory", "researcher")
+Task("Coder", "Implement features with memory coordination", "coder")
+Task("Tester", "Create tests and report results", "tester")
+Task("Reviewer", "Review code quality and share feedback", "reviewer")
```
+## Available Agent Types
+
+### Swarm Agents
+- **hierarchical-coordinator** - Tree-structure command coordination
+- **mesh-coordinator** - Peer-to-peer collaboration
+- **adaptive-coordinator** - Dynamic topology adjustment
+
+### Hive-Mind Agents (NEW)
+- **queen-coordinator** - Sovereign orchestrator
+- **collective-intelligence-coordinator** - Distributed cognition
+- **swarm-memory-manager** - Memory coordination
+- **worker-specialist** - Task execution
+- **scout-explorer** - Reconnaissance
+
+### Core Development
+- **coder**, **reviewer**, **tester**, **planner**, **researcher**
+
+### Specialized Agents
+- **backend-dev**, **mobile-dev**, **ml-developer**
+- **api-docs**, **system-architect**, **code-analyzer**
+- **base-template-generator**, **cicd-engineer**
+
## MCP Coordination Setup (Optional)
-MCP tools are ONLY for coordination:
+MCP tools are ONLY for coordination topology:
```javascript
-mcp__claude-flow__swarm_init { topology: "mesh" }
+// Step 1: Initialize swarm topology
+mcp__claude-flow__swarm_init {
+ topology: "hierarchical",
+ maxAgents: 10
+}
+
+// Step 2: Define agent types for coordination
mcp__claude-flow__agent_spawn { type: "researcher" }
+mcp__claude-flow__agent_spawn { type: "coder" }
+```
+
+## Memory Coordination Protocol
+
+All spawned agents MUST write to memory:
+
+```javascript
+// MANDATORY pattern for every agent
+mcp__claude-flow__memory_usage {
+ action: "store",
+ key: "swarm/[agent-name]/status",
+ namespace: "coordination",
+ value: JSON.stringify({
+ agent: "agent-name",
+ status: "active",
+ task: "current task",
+ timestamp: Date.now()
+ })
+}
```
## Best Practices
-1. Always spawn agents concurrently
-2. Use Task tool for execution
-3. MCP only for coordination
-4. Batch all operations
+
+1. **Always spawn agents concurrently** - Use ONE message
+2. **Use Task tool for execution** - Claude Code handles real work
+3. **MCP only for coordination** - Setup topology, not execution
+4. **Batch all operations** - 300% performance improvement
+5. **Memory writes are MANDATORY** - Agents must coordinate
+
+## Example: Full Stack Development
+
+```javascript
+// Single message with ALL agents
+[Concurrent Execution]:
+ Task("Backend Developer", "Build REST API with memory coordination", "backend-dev")
+ Task("Frontend Developer", "Create React UI, check memory for API", "coder")
+ Task("Database Architect", "Design schema, store in memory", "code-analyzer")
+ Task("Test Engineer", "Write tests, check memory for contracts", "tester")
+ Task("DevOps Engineer", "Setup CI/CD, document in memory", "cicd-engineer")
+
+ TodoWrite { todos: [...all todos...] }
+
+ Write "backend/server.js"
+ Write "frontend/App.jsx"
+ Write "database/schema.sql"
+```
+
+## See Also
+- `swarm-init` - Initialize swarm topology
+- `hive-mind-spawn` - Spawn hive-mind agents
+- `task-orchestrate` - Orchestrate complex tasks
+- `agent-types` - List all available agents