aidevops 2.52.1

package/.agent/tools/ai-orchestration/langflow.md

@@ -0,0 +1,405 @@
+---
+description: Langflow visual AI workflow builder - setup, usage, and integration
+mode: subagent
+tools:
+  read: true
+  write: true
+  edit: true
+  bash: true
+  glob: true
+  grep: true
+  webfetch: true
+---
+
+# Langflow - Visual AI Workflow Builder
+
+<!-- AI-CONTEXT-START -->
+
+## Quick Reference
+
+- **Purpose**: Visual drag-and-drop builder for AI-powered agents and workflows
+- **License**: MIT (fully open-source, commercial use permitted)
+- **Setup**: `bash .agent/scripts/langflow-helper.sh setup`
+- **Start**: `~/.aidevops/scripts/start-langflow.sh`
+- **Stop**: `~/.aidevops/scripts/stop-langflow.sh`
+- **Status**: `~/.aidevops/scripts/langflow-status.sh`
+- **URL**: http://localhost:7860
+- **Config**: `~/.aidevops/langflow/.env`
+- **Install**: `pip install langflow` in venv at `~/.aidevops/langflow/venv/`
+- **Privacy**: Flows stored locally, optional cloud sync
+
+**Key Features**:
+
+- Drag-and-drop visual flow builder
+- Export flows to Python code
+- Built-in MCP server support
+- LangChain ecosystem integration
+- Local LLM support (Ollama)
+
+<!-- AI-CONTEXT-END -->
+
+## Overview
+
+Langflow is a powerful visual tool for building and deploying AI-powered agents and workflows. It provides developers with both a visual authoring experience and built-in API/MCP servers that turn every workflow into a tool.
+
+## Installation
+
+### Automated Setup (Recommended)
+
+```bash
+# Run the setup script
+bash .agent/scripts/langflow-helper.sh setup
+
+# Configure API keys
+nano ~/.aidevops/langflow/.env
+
+# Start Langflow
+~/.aidevops/scripts/start-langflow.sh
+```
+
+### Manual Installation
+
+```bash
+# Create directory and virtual environment
+mkdir -p ~/.aidevops/langflow
+cd ~/.aidevops/langflow
+python3 -m venv venv
+source venv/bin/activate
+
+# Install Langflow
+pip install langflow
+
+# Run Langflow
+langflow run
+```
+
+### Docker Installation
+
+```bash
+# Run with Docker
+docker run -p 7860:7860 langflowai/langflow:latest
+
+# With persistent storage
+docker run -p 7860:7860 -v langflow_data:/app/langflow langflowai/langflow:latest
+```
+
+### Desktop App
+
+Download Langflow Desktop from https://www.langflow.org/desktop for Windows/macOS.
+
+## Configuration
+
+### Environment Variables
+
+Create `~/.aidevops/langflow/.env`:
+
+```bash
+# OpenAI Configuration
+OPENAI_API_KEY=your_openai_api_key_here
+
+# Anthropic Configuration (optional)
+ANTHROPIC_API_KEY=your_anthropic_key_here
+
+# Langflow Configuration
+LANGFLOW_HOST=0.0.0.0
+LANGFLOW_PORT=7860
+LANGFLOW_WORKERS=1
+
+# Database (default: SQLite)
+LANGFLOW_DATABASE_URL=sqlite:///./langflow.db
+
+# Local LLM (Ollama)
+OLLAMA_BASE_URL=http://localhost:11434
+```
+
+### Custom Components
+
+Create custom components in Python:
+
+```python
+# ~/.aidevops/langflow/components/my_component.py
+from langflow.custom import CustomComponent
+from langflow.schema import Data
+
+class MyCustomComponent(CustomComponent):
+    display_name = "My Custom Component"
+    description = "A custom component for aidevops"
+    
+    def build(self, input_text: str) -> Data:
+        # Your custom logic here
+        result = input_text.upper()
+        return Data(text=result)
+```
+
+Load custom components:
+
+```bash
+langflow run --components-path ~/.aidevops/langflow/components/
+```
+
+## Usage
+
+### Starting Services
+
+```bash
+# Start Langflow
+~/.aidevops/scripts/start-langflow.sh
+
+# Check status
+~/.aidevops/scripts/langflow-status.sh
+
+# Stop Langflow
+~/.aidevops/scripts/stop-langflow.sh
+```
+
+### Accessing the Interface
+
+- **Web UI**: http://localhost:7860
+- **API Docs**: http://localhost:7860/docs
+- **Health Check**: http://localhost:7860/health
+
+### Building Your First Flow
+
+1. Open http://localhost:7860
+2. Click "New Flow" or use a template
+3. Drag components from the sidebar
+4. Connect components by dragging edges
+5. Configure each component's parameters
+6. Click "Run" to test the flow
+
+### Common Flow Patterns
+
+**RAG Pipeline**:
+
+```text
+[Document Loader] → [Text Splitter] → [Embeddings] → [Vector Store]
+                                                           ↓
+[User Input] → [Retriever] → [LLM] → [Output]
+```
+
+**Multi-Agent Chat**:
+
+```text
+[User Input] → [Router Agent] → [Specialist Agent 1]
+                             → [Specialist Agent 2]
+                             → [Aggregator] → [Output]
+```
+
+## API Integration
+
+### REST API
+
+```python
+import requests
+
+# Run a flow
+response = requests.post(
+    "http://localhost:7860/api/v1/run/<flow-id>",
+    json={
+        "input_value": "Hello, world!",
+        "output_type": "chat",
+        "input_type": "chat"
+    }
+)
+print(response.json())
+```
+
+### MCP Server
+
+Langflow can expose flows as MCP tools:
+
+```bash
+# Start with MCP server enabled
+langflow run --mcp
+
+# Or configure in .env
+LANGFLOW_MCP_ENABLED=true
+```
+
+Then connect from AI assistants that support MCP.
+
+## Git Integration
+
+### Exporting Flows
+
+```bash
+# Export a flow to JSON
+langflow export --flow-id <flow-id> --output flows/my-flow.json
+
+# Export all flows
+langflow export --all --output flows/
+```
+
+### Importing Flows
+
+```bash
+# Import a flow from JSON
+langflow import --file flows/my-flow.json
+
+# Import all flows from directory
+langflow import --directory flows/
+```
+
+### Version Control Best Practices
+
+```bash
+# .gitignore additions
+langflow.db
+*.log
+__pycache__/
+.env
+
+# Track these
+flows/*.json
+components/*.py
+```
+
+### Bi-directional Sync
+
+Use file watchers for automatic sync:
+
+```python
+# sync_flows.py
+from watchdog.observers import Observer
+from watchdog.events import FileSystemEventHandler
+import subprocess
+
+class FlowSyncHandler(FileSystemEventHandler):
+    def on_modified(self, event):
+        if event.src_path.endswith('.json'):
+            subprocess.run(['langflow', 'import', '--file', event.src_path])
+
+# Run with: python sync_flows.py
+```
+
+## Local LLM Support
+
+### Using Ollama
+
+```bash
+# Install Ollama
+curl -fsSL https://ollama.com/install.sh | sh
+
+# Pull a model
+ollama pull llama3.2
+ollama pull codellama
+
+# Configure in Langflow
+# Add Ollama component and set base URL to http://localhost:11434
+```
+
+### Using LM Studio
+
+1. Download LM Studio from https://lmstudio.ai
+2. Load a model and start the local server
+3. In Langflow, use OpenAI-compatible endpoint: http://localhost:1234/v1
+
+## Deployment
+
+### Docker Compose
+
+```yaml
+# docker-compose.yml
+version: '3.8'
+services:
+  langflow:
+    image: langflowai/langflow:latest
+    ports:
+      - "7860:7860"
+    volumes:
+      - langflow_data:/app/langflow
+      - ./flows:/app/flows
+    environment:
+      - OPENAI_API_KEY=${OPENAI_API_KEY}
+    restart: unless-stopped
+
+volumes:
+  langflow_data:
+```
+
+### Production Considerations
+
+- Use PostgreSQL instead of SQLite for production
+- Enable authentication for multi-user deployments
+- Use reverse proxy (nginx/traefik) for HTTPS
+- Set up monitoring and logging
+
+## Troubleshooting
+
+### Common Issues
+
+**Port already in use**:
+
+```bash
+# Find and kill process
+lsof -i :7860
+kill -9 <PID>
+```
+
+**Database errors**:
+
+```bash
+# Reset database
+rm ~/.aidevops/langflow/langflow.db
+langflow run
+```
+
+**Component not loading**:
+
+```bash
+# Check component syntax
+python -c "from components.my_component import MyCustomComponent"
+```
+
+### Logs
+
+```bash
+# View logs
+tail -f ~/.aidevops/langflow/langflow.log
+
+# Debug mode
+LANGFLOW_LOG_LEVEL=DEBUG langflow run
+```
+
+## Integration Examples
+
+### With aidevops Workflows
+
+```bash
+# Export flow for version control
+langflow export --flow-id <id> --output flows/devops-automation.json
+
+# Commit to git
+git add flows/devops-automation.json
+git commit -m "feat: add DevOps automation flow"
+```
+
+### With CrewAI
+
+Langflow can orchestrate CrewAI agents:
+
+1. Create a custom component that imports CrewAI
+2. Define agents and tasks in the component
+3. Connect to other Langflow components
+
+### With OpenCode
+
+Use Langflow flows as tools in OpenCode via MCP:
+
+```json
+{
+  "mcpServers": {
+    "langflow": {
+      "command": "langflow",
+      "args": ["run", "--mcp"]
+    }
+  }
+}
+```
+
+## Resources
+
+- **Documentation**: https://docs.langflow.org
+- **GitHub**: https://github.com/langflow-ai/langflow
+- **Discord**: https://discord.gg/EqksyE2EX9
+- **Templates**: https://www.langflow.org/templates