@bluefly/openstandardagents 0.3.6 → 0.4.1

package/examples/export/langchain/production-agent-with-memory/README.md

@@ -0,0 +1,373 @@
+# Production Agent with Memory - LangChain Export (v0.4.1)
+
+**Showcase Example for OSSA v0.4.1 LangChain Memory Features**
+
+This example demonstrates all v0.4.1 production-quality memory features for LangChain exports:
+
+## ✨ Features Demonstrated
+
+### 1. **Conversation Buffer Memory** 💬
+- Window management (configurable message history)
+- Memory statistics and monitoring
+- Structured logging
+- Production-ready error handling
+
+### 2. **Conversation Summary Memory** 📝
+- Automatic summarization when token limit reached
+- Configurable token limits
+- Summary regeneration
+- Cost-effective model (gpt-4o-mini) for summarization
+
+### 3. **Entity Memory** 🏷️
+- Automatic entity extraction and tracking
+- Entity context retrieval
+- Relationship mapping
+- Entity listing and management
+
+### 4. **Redis Persistence** 🔴
+- Connection pooling with automatic retry
+- Exponential backoff on connection failures
+- Health checks and monitoring
+- Session management (list, delete, get stats)
+- TTL configuration per session
+
+### 5. **PostgreSQL Persistence** 🐘
+- Connection pooling (ThreadedConnectionPool)
+- Automatic schema initialization
+- Session analytics and statistics
+- Export functionality (JSON, CSV)
+- Health checks and monitoring
+- Long-term storage for compliance
+
+## 🎯 Memory Backends
+
+| Backend | Type | Persistence | Best For |
+|---------|------|-------------|----------|
+| **buffer** | In-Memory | No | Development, short conversations |
+| **summary** | In-Memory | No | Long conversations, token efficiency |
+| **entity** | In-Memory | No | Entity-focused interactions |
+| **redis** | Persistent | Yes | Production, multi-instance, session persistence |
+| **postgres** | Persistent | Yes | Analytics, compliance, long-term storage |
+
+## 📋 Export Command
+
+```bash
+# Export with different memory backends
+
+# Buffer memory (in-memory)
+ossa export agent.ossa.yaml -p langchain -o ./buffer-agent
+# Memory generated: memory.py with ConversationBufferWindowMemory
+
+# Summary memory (token-efficient)
+ossa export agent.ossa.yaml -p langchain --memory-backend summary -o ./summary-agent
+# Memory generated: memory.py with ConversationSummaryBufferMemory
+
+# Entity memory (entity tracking)
+ossa export agent.ossa.yaml -p langchain --memory-backend entity -o ./entity-agent
+# Memory generated: memory.py with ConversationEntityMemory
+
+# Redis persistence (production)
+ossa export agent.ossa.yaml -p langchain --memory-backend redis -o ./redis-agent
+# Memory generated: memory.py with Redis-backed persistence
+
+# PostgreSQL persistence (analytics)
+ossa export agent.ossa.yaml -p langchain --memory-backend postgres -o ./postgres-agent
+# Memory generated: memory.py with PostgreSQL-backed persistence
+```
+
+## 📦 Generated Memory Features
+
+All memory backends include:
+- ✅ **get_memory()** - Create memory instance
+- ✅ **clear_memory()** - Clear conversation history
+- ✅ **get_memory_stats()** - Get usage statistics
+- ✅ **Logging** - Structured logging with context
+- ✅ **Error Handling** - Comprehensive try/except with exc_info
+
+Persistent backends (Redis, PostgreSQL) additionally include:
+- ✅ **validate_connection()** - Connection validation with retry
+- ✅ **health_check()** - Health monitoring
+- ✅ **get_all_sessions()** - List all sessions
+- ✅ **delete_session()** - Delete specific session
+
+## 🎯 Configuration Options
+
+### OSSA Manifest Configuration
+
+```yaml
+spec:
+  memory:
+    enabled: true
+    type: conversation_buffer  # or: conversation_summary, entity
+    window_size: 20            # Number of messages to keep
+    max_token_limit: 4000      # Token limit for summary
+    return_messages: true      # Return as Message objects
+
+    # Persistence (optional)
+    persistence:
+      enabled: true
+      backend: redis           # or: postgres
+      ttl: 86400              # Time-to-live (seconds)
+      connection:
+        url: ${REDIS_URL}     # Connection URL from env
+        pool_size: 10         # Connection pool size
+        timeout: 30           # Connection timeout (seconds)
+```
+
+## 📊 Code Examples
+
+### Buffer Memory with Window
+
+```python
+import logging
+from memory import get_memory, clear_memory, get_memory_stats
+
+logger = logging.getLogger(__name__)
+
+# Create memory (keeps last 20 messages)
+memory = get_memory()
+
+# Use with agent
+agent = create_agent(memory=memory)
+response = agent.run("Hello!")
+
+# Get statistics
+stats = get_memory_stats(memory)
+# {
+#   "type": "buffer_window",
+#   "window_size": 20,
+#   "message_count": 5,
+#   "messages": [...]
+# }
+
+# Clear when done
+clear_memory(memory)
+```
+
+### Summary Memory
+
+```python
+from memory import get_memory, regenerate_summary, get_memory_stats
+
+# Create summary memory (auto-summarizes at 4000 tokens)
+memory = get_memory()
+
+# Use with agent
+agent = create_agent(memory=memory)
+
+# Get statistics
+stats = get_memory_stats(memory)
+# {
+#   "type": "summary_buffer",
+#   "max_token_limit": 4000,
+#   "message_count": 10,
+#   "has_summary": true
+# }
+
+# Force summary regeneration
+summary = regenerate_summary(memory)
+print(f"Summary: {summary}")
+```
+
+### Entity Memory
+
+```python
+from memory import (
+    get_memory,
+    get_entity_context,
+    list_entities,
+    get_memory_stats
+)
+
+# Create entity memory
+memory = get_memory()
+
+# Use with agent
+agent = create_agent(memory=memory)
+agent.run("John Smith is the CEO of TechCorp")
+
+# List tracked entities
+entities = list_entities(memory)
+# ["John Smith", "TechCorp"]
+
+# Get entity context
+context = get_entity_context(memory, "John Smith")
+# "John Smith is the CEO of TechCorp"
+
+# Get statistics
+stats = get_memory_stats(memory)
+# {
+#   "type": "entity",
+#   "entity_count": 2,
+#   "entities": ["John Smith", "TechCorp"],
+#   "entity_details": {...}
+# }
+```
+
+### Redis Persistence
+
+```python
+from memory import (
+    get_memory,
+    validate_redis_connection,
+    health_check,
+    get_all_sessions,
+    delete_session,
+    get_memory_stats
+)
+
+# Validate connection first (with retry)
+if not validate_redis_connection(max_retries=3):
+    raise ConnectionError("Failed to connect to Redis")
+
+# Create memory for session
+memory = get_memory(session_id="user-123")
+
+# Use with agent
+agent = create_agent(memory=memory)
+
+# Get session statistics
+stats = get_memory_stats(session_id="user-123")
+# {
+#   "type": "redis",
+#   "session_id": "user-123",
+#   "message_count": 42,
+#   "ttl_remaining": 82800,  # seconds
+#   "key": "langchain:chat:user-123"
+# }
+
+# Health check
+health = health_check()
+# {
+#   "healthy": true,
+#   "latency_ms": 2.5,
+#   "redis_version": "7.0.0",
+#   "uptime_seconds": 86400
+# }
+
+# List all sessions
+sessions = get_all_sessions()
+# ["user-123", "user-456", ...]
+
+# Delete session
+success = delete_session("user-123")
+```
+
+### PostgreSQL Persistence
+
+```python
+from memory import (
+    get_memory,
+    validate_postgres_connection,
+    initialize_schema,
+    health_check,
+    get_all_sessions,
+    export_session_history,
+    get_memory_stats
+)
+
+# Validate connection and initialize schema
+if not validate_postgres_connection(max_retries=3):
+    raise ConnectionError("Failed to connect to PostgreSQL")
+
+initialize_schema()
+
+# Create memory for session
+memory = get_memory(session_id="user-123")
+
+# Use with agent
+agent = create_agent(memory=memory)
+
+# Get session statistics
+stats = get_memory_stats(session_id="user-123")
+# {
+#   "type": "postgres",
+#   "session_id": "user-123",
+#   "message_count": 42,
+#   "first_message_at": "2026-02-03T10:00:00",
+#   "last_message_at": "2026-02-03T11:30:00"
+# }
+
+# Health check
+health = health_check()
+# {
+#   "healthy": true,
+#   "latency_ms": 5.2,
+#   "postgres_version": "PostgreSQL 15.1",
+#   "schema_initialized": true
+# }
+
+# Export session history
+messages_json = export_session_history("user-123", format="json")
+messages_csv = export_session_history("user-123", format="csv")
+```
+
+## 🧪 Testing
+
+All features are tested with 18 comprehensive unit tests:
+
+```bash
+npm test tests/unit/services/export/langchain/memory-generator.test.ts
+```
+
+**Test Coverage:**
+- ✅ Configuration parsing (3 tests)
+- ✅ Buffer memory (2 tests)
+- ✅ Summary memory (2 tests)
+- ✅ Entity memory (2 tests)
+- ✅ Redis persistence (3 tests)
+- ✅ PostgreSQL persistence (3 tests)
+- ✅ Production features (3 tests)
+
+## 🔧 Environment Variables
+
+**Redis Backend:**
+```bash
+REDIS_URL=redis://localhost:6379
+```
+
+**PostgreSQL Backend:**
+```bash
+POSTGRES_URL=postgresql://postgres:postgres@localhost:5432/agent_memory
+```
+
+**OpenAI API (for Summary/Entity Memory):**
+```bash
+OPENAI_API_KEY=sk-...
+```
+
+## 📊 Version Information
+
+- **OSSA Version**: v0.4.1+
+- **Release**: v0.4.1 (Feb 14, 2026)
+- **Feature Set**: LangChain Memory Support
+- **Status**: ✅ Complete
+
+## 🎓 What's Next
+
+This is the **Week 2 deliverable** of v0.4.1 (part 1). Next feature:
+
+- **Week 2 (part 2)**: Streaming Support (SSE, WebSocket, a2a integration)
+
+## 📚 Documentation
+
+- [v0.4.1-v0.4.8 Release Plan](../../../../../wikis/technical-docs.wiki/action-items/Ossa-PLAN/v0.4.1-v0.4.8-release-plan.md)
+- [LangChain Export Guide](../../../../../docs/export/langchain.md)
+- [Memory Configuration Guide](../../../../../docs/memory/configuration.md)
+- [OSSA Specification](https://openstandardagents.org/spec)
+
+## 🤝 Contributing
+
+This example demonstrates production quality standards for memory implementations:
+
+1. **Connection Management**: Use connection pooling
+2. **Error Handling**: Comprehensive try/except with logging
+3. **Retry Logic**: Exponential backoff for failed connections
+4. **Health Checks**: Monitoring and diagnostics
+5. **Statistics**: Usage tracking and analytics
+
+---
+
+**Generated by**: OSSA v0.4.1 LangChain Memory Generator
+**Date**: 2026-02-03
+**Status**: Production Ready ✅

package/examples/export/langchain/production-agent-with-memory/agent.ossa.yaml

@@ -0,0 +1,97 @@
+apiVersion: ossa/v0.4.1
+kind: Agent
+metadata:
+  name: production-agent-with-memory
+  version: 1.0.0
+  description: |
+    Production-ready agent with advanced memory features showcasing v0.4.1:
+    - Conversation buffer with window management
+    - Conversation summary with token limits
+    - Entity memory for entity tracking
+    - Redis persistence for multi-instance deployments
+    - PostgreSQL persistence for analytics and compliance
+  labels:
+    platform: langchain
+    export: langchain
+    use-case: production-showcase
+    tier: production
+    version: v0.4.1
+
+spec:
+  role: |
+    You are a production assistant with advanced memory capabilities.
+    You maintain conversation context across sessions and track entities.
+
+  llm:
+    provider: openai
+    model: gpt-4o
+    temperature: 0.7
+    maxTokens: 2000
+
+  # Memory Configuration (Full OSSA spec)
+  memory:
+    enabled: true
+    type: conversation_buffer
+    window_size: 20
+    max_token_limit: 4000
+    return_messages: true
+
+    # Persistence Configuration
+    persistence:
+      enabled: true
+      backend: redis  # or: postgres
+      ttl: 86400  # 24 hours
+      connection:
+        url: ${REDIS_URL}  # or ${POSTGRES_URL}
+        pool_size: 10
+        timeout: 30
+
+  tools:
+    - type: function
+      name: remember_fact
+      description: Remember an important fact
+      parameters:
+        type: object
+        properties:
+          fact:
+            type: string
+            description: Fact to remember
+        required:
+          - fact
+
+    - type: function
+      name: recall_conversation
+      description: Recall previous conversation
+      parameters:
+        type: object
+        properties:
+          query:
+            type: string
+            description: What to recall
+        required:
+          - query
+
+  autonomy:
+    level: supervised
+    approval_required: false
+
+  observability:
+    logging:
+      level: info
+      format: json
+    metrics:
+      enabled: true
+    tracing:
+      enabled: true
+      provider: opentelemetry
+
+extensions:
+  langchain:
+    agent_type: react
+    memory_type: conversation_buffer  # buffer, summary, entity
+    export_config:
+      format: python
+      include_dependencies: true
+      include_api_server: true
+      api_port: 8000
+      memory_backend: redis  # buffer, summary, entity, redis, postgres