/add-redisvl

Arguments: $ARGUMENTS

Goal: Integrate RedisVL for vector similarity search with HNSW indexing

Core Principles:

• Detect project language and framework before setup
• Use environment variables for Redis configuration
• Follow RedisVL best practices for vector search
• Provide clear setup instructions and examples

Phase 1: Discovery Goal: Understand project structure and requirements

Actions:

• Parse $ARGUMENTS for project path (default to current directory)
• Detect project language (Python, Node.js, etc.)
• Check if Redis is already configured
• Load package manager files: @package.json or @pyproject.toml or @requirements.txt
• Detect existing vector search implementations

Phase 2: Requirements Gathering Goal: Clarify vector search configuration needs

Actions:

• If requirements are unclear, use AskUserQuestion to gather:
  • What data will be indexed? (documents, images, embeddings)
  • What embedding model? (OpenAI, Cohere, local model)
  • What distance metric? (cosine, euclidean, dot product)
  • Index size expectations? (1K, 100K, 1M+ vectors)
  • HNSW parameters? (M, ef_construction, ef_runtime)
• Document user requirements for agent context

Phase 3: Pre-flight Validation Goal: Verify environment readiness

Actions:

• Check Redis availability: !{bash redis-cli ping 2>/dev/null || echo "Redis not running"}
• Verify Redis version supports vector search (7.2+)
• Check if RedisVL is already installed
• Detect conflicts with existing vector search libraries

Phase 4: Implementation Goal: Execute RedisVL integration

Actions:

Task(description="Add RedisVL integration", subagent_type="redis:redisvl-integrator", prompt="You are the redisvl-integrator agent. Add RedisVL (Redis Vector Library) integration for $ARGUMENTS.

Project Context:

• Language: [Detected from Phase 1]
• Framework: [Detected from Phase 1]
• Existing Redis config: [From Phase 1]

User Requirements:

• Data type: [From Phase 2]
• Embedding model: [From Phase 2]
• Distance metric: [From Phase 2]
• Index size: [From Phase 2]
• HNSW parameters: [From Phase 2]

Implementation Tasks:

1. Install RedisVL package (pip install redisvl OR npm install redisvl)
2. Configure Redis connection with environment variables
3. Create vector index schema with HNSW algorithm
4. Generate example code for:
   • Creating vector index
   • Storing embeddings
   • Performing similarity search
   • Updating/deleting vectors
5. Add configuration file (.env.example with placeholders)
6. Create setup documentation

Security Requirements:

• Use REDIS_URL environment variable
• Never hardcode credentials
• Add .env to .gitignore
• Document where to get Redis credentials

Expected Deliverables:

• RedisVL package installed
• Vector index schema file
• Example usage code
• Environment variable template
• Setup documentation with next steps")

Phase 5: Verification Goal: Validate RedisVL integration

Actions:

• Verify RedisVL package is installed
• Check environment variable template exists
• Validate example code syntax
• Test Redis connection if available
• Run type checking if TypeScript: !{bash npm run typecheck 2>/dev/null || true}
• Run linting if configured: !{bash npm run lint 2>/dev/null || pylint . 2>/dev/null || true}

Phase 6: Summary Goal: Document what was accomplished

Actions:

• Summarize changes made:
  • Package installed
  • Files created/modified
  • Vector index configuration
  • HNSW parameters used
• Provide next steps:
  • Set Redis URL in environment
  • Configure embedding model
  • Run example code to test
  • Customize HNSW parameters for production
• Share relevant file paths (absolute paths)
• Display RedisVL documentation links