claude-flow-novice 1.3.2 → 1.3.3

package/.claude/commands/coordination/coordination-system.md

@@ -0,0 +1,88 @@
+# Agent Coordination System
+
+## Overview
+The Claude-Flow coordination system manages multiple AI agents working together on complex tasks. It provides intelligent task distribution, resource management, and inter-agent communication.
+
+## Agent Types and Capabilities
+- **Researcher**: Web search, information gathering, knowledge synthesis
+- **Coder**: Code analysis, development, debugging, testing
+- **Analyst**: Data processing, pattern recognition, insights generation
+- **Coordinator**: Task planning, resource allocation, workflow management
+- **General**: Multi-purpose agent with balanced capabilities
+
+## Task Management
+- **Priority Levels**: 1 (lowest) to 10 (highest)
+- **Dependencies**: Tasks can depend on completion of other tasks
+- **Parallel Execution**: Independent tasks run concurrently
+- **Load Balancing**: Automatic distribution based on agent capacity
+
+## Coordination Commands
+```bash
+# Agent Management
+npx claude-flow agent spawn <type> --name <name> --priority <1-10>
+npx claude-flow agent list
+npx claude-flow agent info <agent-id>
+npx claude-flow agent terminate <agent-id>
+
+# Task Management  
+npx claude-flow task create <type> <description> --priority <1-10> --deps <task-ids>
+npx claude-flow task list --verbose
+npx claude-flow task status <task-id>
+npx claude-flow task cancel <task-id>
+
+# System Monitoring
+npx claude-flow status --verbose
+npx claude-flow monitor --interval 5000
+```
+
+## Workflow Execution
+Workflows are defined in JSON format and can orchestrate complex multi-agent operations:
+```bash
+npx claude-flow workflow examples/research-workflow.json
+npx claude-flow workflow examples/development-config.json --async
+```
+
+## Advanced Features
+- **Circuit Breakers**: Automatic failure handling and recovery
+- **Work Stealing**: Dynamic load redistribution for efficiency
+- **Resource Limits**: Memory and CPU usage constraints
+- **Metrics Collection**: Performance monitoring and optimization
+
+## Configuration
+Coordination settings in `claude-flow.config.json`:
+```json
+{
+  "orchestrator": {
+    "maxConcurrentTasks": 10,
+    "taskTimeout": 300000,
+    "defaultPriority": 5
+  },
+  "agents": {
+    "maxAgents": 20,
+    "defaultCapabilities": ["research", "code", "terminal"],
+    "resourceLimits": {
+      "memory": "1GB",
+      "cpu": "50%"
+    }
+  }
+}
+```
+
+## Communication Patterns
+- **Direct Messaging**: Agent-to-agent communication
+- **Event Broadcasting**: System-wide notifications
+- **Shared Memory**: Common information access
+- **Task Handoff**: Seamless work transfer between agents
+
+## Best Practices
+- Start with general agents and specialize as needed
+- Use descriptive task names and clear requirements
+- Monitor system resources during heavy workloads
+- Implement proper error handling in workflows
+- Regular cleanup of completed tasks and inactive agents
+
+## Troubleshooting
+- Check agent health with `npx claude-flow status`
+- View detailed logs with `npx claude-flow monitor`
+- Restart stuck agents with terminate/spawn cycle
+- Use `--verbose` flags for detailed diagnostic information

package/.claude/commands/memory/memory-bank.md

@@ -0,0 +1,58 @@
+# Memory Bank Configuration
+
+## Overview
+The Claude-Flow memory system provides persistent storage and intelligent retrieval of information across agent sessions. It uses a hybrid approach combining SQL databases with semantic search capabilities.
+
+## Storage Backends
+- **Primary**: JSON database (`./memory/claude-flow-data.json`)
+- **Sessions**: File-based storage in `./memory/sessions/`
+- **Cache**: In-memory cache for frequently accessed data
+
+## Memory Organization
+- **Namespaces**: Logical groupings of related information
+- **Sessions**: Time-bound conversation contexts
+- **Indexing**: Automatic content indexing for fast retrieval
+- **Replication**: Optional distributed storage support
+
+## Commands
+- `npx claude-flow memory query <search>`: Search stored information
+- `npx claude-flow memory stats`: Show memory usage statistics
+- `npx claude-flow memory export <file>`: Export memory to file
+- `npx claude-flow memory import <file>`: Import memory from file
+
+## Configuration
+Memory settings are configured in `claude-flow.config.json`:
+```json
+{
+  "memory": {
+    "backend": "json",
+    "path": "./memory/claude-flow-data.json",
+    "cacheSize": 1000,
+    "indexing": true,
+    "namespaces": ["default", "agents", "tasks", "sessions"],
+    "retentionPolicy": {
+      "sessions": "30d",
+      "tasks": "90d",
+      "agents": "permanent"
+    }
+  }
+}
+```
+
+## Best Practices
+- Use descriptive namespaces for different data types
+- Regular memory exports for backup purposes
+- Monitor memory usage with stats command
+- Clean up old sessions periodically
+
+## Memory Types
+- **Episodic**: Conversation and interaction history
+- **Semantic**: Factual knowledge and relationships
+- **Procedural**: Task patterns and workflows
+- **Meta**: System configuration and preferences
+
+## Integration Notes
+- Memory is automatically synchronized across agents
+- Search supports both exact match and semantic similarity
+- Memory contents are private to your local instance
+- No data is sent to external services without explicit commands

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "claude-flow-novice",
-  "version": "1.3.2",
+  "version": "1.3.3",
   "description": "Standalone Claude Flow for beginners - AI agent orchestration made easy. Enhanced init command creates complete agent system, MCP configuration with 30 essential tools, and automated hooks. Fully standalone with zero external dependencies, all memory leaks eliminated, complete project setup in one command.",
   "mcpName": "io.github.ruvnet/claude-flow",
   "main": ".claude-flow-novice/dist/index.js",