/integrate-langchain

Arguments: $ARGUMENTS Security: @~/.claude/plugins/marketplaces/dev-lifecycle-marketplace/docs/security/SECURITY-RULES.md

Goal: Integrate LangChain vector store and memory capabilities with Redis for AI applications

Core Principles:

• Detect existing LangChain usage before adding
• Ask user which features they need (vector store, memory, chat history, semantic cache)
• Use RedisVectorStore for embeddings and semantic search
• Implement appropriate memory types based on use case
• Follow LangChain best practices and patterns

Phase 1: Discovery Goal: Understand project context and LangChain usage

Actions:

• Detect if LangChain is already installed
• Check for existing vector store or memory implementations
• Identify AI/LLM usage patterns in codebase
• Example: !{bash grep -r "langchain" . --include=".py" --include=".js" --include="*.ts" 2>/dev/null | head -5}

Phase 2: Requirements Gathering Goal: Determine which LangChain + Redis features to integrate

Actions:

• Use AskUserQuestion to determine needs:
  • Which features? (vector store, conversation memory, entity memory, summary memory, chat history, semantic cache)
  • What embedding model? (OpenAI, Anthropic, local model)
  • Memory retention policy? (session-based, persistent, TTL)
  • Semantic similarity threshold? (0.85-0.95 recommended)
• Confirm project type (Python, JavaScript/TypeScript)
• Validate Redis connection details available

Phase 3: Planning Goal: Design the LangChain + Redis integration approach

Actions:

• Based on requirements, plan integration:
  • Vector store: RedisVectorStore with HNSW indexes
  • Memory types: ConversationBufferMemory, EntityMemory, ConversationSummaryMemory
  • Chat history: RedisChatMessageHistory for persistence
  • Semantic cache: Integrated with LangChain cache layer
• Identify files to create/modify
• Present plan to user for approval

Phase 4: Implementation Goal: Integrate LangChain features with Redis

Actions:

Task(description="Integrate LangChain with Redis", subagent_type="redis:langchain-integrator", prompt="You are the redis:langchain-integrator agent. Integrate LangChain vector store and memory with Redis for $ARGUMENTS.

Context: User selected the following features to integrate Features requested: [List from Phase 2] Embedding model: [From Phase 2] Memory policy: [From Phase 2] Similarity threshold: [From Phase 2]

Requirements:

• Install langchain-redis package and dependencies
• Configure RedisVectorStore with HNSW indexes for vector similarity
• Implement selected memory types (conversation, entity, summary)
• Set up RedisChatMessageHistory for chat persistence
• Integrate semantic cache if requested
• Follow LangChain patterns for retrieval chains
• Add environment variable configuration (REDIS_URL, embedding API keys)
• Create example usage code showing integration
• Add error handling and connection validation
• Include cost optimization recommendations

Expected output:

• LangChain + Redis integration code
• Configuration files with placeholder API keys
• Example usage demonstrating features
• Documentation on memory patterns and retention policies")

Phase 5: Verification Goal: Validate LangChain + Redis integration works

Actions:

• Verify all dependencies installed
• Check configuration files use placeholders (no hardcoded keys)
• Test vector store connection and indexing
• Test memory persistence and retrieval
• Validate chat history storage
• Example: !{bash python -c "from langchain_redis import RedisVectorStore; print('Import successful')" 2>&1}

Phase 6: Summary Goal: Document integration and provide guidance

Actions:

• Summarize what was integrated:
  • Vector store capabilities (embedding search, similarity)
  • Memory types implemented (conversation, entity, summary)
  • Chat history persistence
  • Semantic cache if enabled
• Highlight key files created/modified
• Provide cost optimization tips:
  • Semantic cache hit rate monitoring
  • Memory TTL tuning
  • Embedding reuse strategies
• Suggest next steps:
  • Test with real queries
  • Tune similarity thresholds
  • Monitor memory usage
  • Set up chat history cleanup policies