/integrate-llamaindex

Arguments: $ARGUMENTS

Goal: Set up complete LlamaIndex integration with Redis vector store for semantic search, RAG applications, and AI-powered data retrieval

Core Principles:

• Detect project structure and framework
• Use RedisVectorStore for optimal performance
• Configure appropriate query and chat engines
• Follow LlamaIndex best practices
• Provide working examples

Phase 1: Discovery Goal: Understand project context and requirements

Actions:

• Parse $ARGUMENTS for integration type (basic, rag, chat, custom)
• Detect project structure and dependencies
• Check for existing LlamaIndex or Redis setup
• Example: !{bash test -f package.json && echo "node" || test -f pyproject.toml && echo "python" || test -f requirements.txt && echo "python"}
• Load existing configuration files if present
• Example: @.env

Phase 2: Requirements Gathering Goal: Clarify integration specifications

Actions:

• If $ARGUMENTS is unclear or minimal, use AskUserQuestion to gather:
  • Which LlamaIndex features? (VectorStoreIndex, query engine, chat engine, agents)
  • Data source types? (documents, text, structured data)
  • Vector embedding model? (OpenAI, HuggingFace, custom)
  • Index strategy? (FLAT, HNSW)
  • Memory requirements? (chat history, conversation memory)
• Summarize requirements and confirm approach

Phase 3: Environment Analysis Goal: Assess current setup and identify gaps

Actions:

• Check for required dependencies
• Python: !{bash pip list | grep -E "llama-index|redis" || echo "Not found"}
• Node.js: !{bash npm list llamaindex redis 2>/dev/null || echo "Not found"}
• Verify Redis connection details
• Check for existing vector indices
• Identify integration points in codebase

Phase 4: Integration Planning Goal: Design the implementation approach

Actions:

• Outline integration strategy based on findings
• Determine file structure (index setup, query handlers, utilities)
• Plan configuration approach (environment variables, config files)
• Identify example use cases to implement
• Present plan to user for confirmation

Phase 5: Implementation Goal: Execute LlamaIndex Redis integration

Actions:

Task(description="Integrate LlamaIndex with Redis", subagent_type="redis:llamaindex-integrator", prompt="You are the llamaindex-integrator agent. Implement complete LlamaIndex integration with Redis for $ARGUMENTS.

Project Context: [Context from Phase 3]

Requirements:

• Set up RedisVectorStore with appropriate configuration
• Implement VectorStoreIndex for semantic search
• Configure query engines (vector, keyword, hybrid)
• Add chat engines with conversation memory if requested
• Create working examples and documentation
• Follow security best practices (no hardcoded API keys)
• Use environment variables for all credentials

Integration Type: [From Phase 2] Embedding Model: [From Phase 2] Index Strategy: [From Phase 2] Additional Features: [From Phase 2]

Expected Output:

• Complete integration code files
• Configuration templates (.env.example)
• Usage examples and documentation
• Test cases or validation scripts")

Phase 6: Verification Goal: Validate the integration works correctly

Actions:

• Review generated integration code
• Check all configuration files use placeholders
• Verify .env.example exists with proper placeholders
• Ensure .gitignore protects secrets
• Run validation if applicable
• Python: !{bash python -m pytest tests/ -v 2>/dev/null || echo "No tests found"}
• Node.js: !{bash npm test 2>/dev/null || echo "No tests configured"}

Phase 7: Summary Goal: Document what was accomplished

Actions:

• Summarize integration components created:
  • Vector store configuration
  • Index setup code
  • Query engine implementations
  • Chat engine setup (if applicable)
  • Example usage files
  • Documentation
• Highlight key configuration options
• Provide next steps:
  • How to obtain API keys (OpenAI, etc.)
  • How to populate vector index with data
  • How to run examples
  • How to customize for specific use cases
• Show file locations with absolute paths