Weaviate

What This Skill Does

The Weaviate skill for OpenClaw provides a robust, production-grade interface for interacting with Weaviate vector databases using the v4 Python client syntax. This skill abstracts the complexities of semantic indexing, hybrid search, and batch data ingestion, ensuring that all interactions adhere to modern best practices. It empowers users to build sophisticated RAG (Retrieval-Augmented Generation) pipelines, configure advanced vectorization modules like text2vec-openai, and manage complex schema configurations without needing to manually write boiler-plate client code.

Installation

To install this skill, run the following command in your terminal: clawhub install openclaw/skills/skills/ivangdavila/weaviate

Use Cases

1. Semantic RAG Engines: Quickly index large document repositories using Weaviate's native vectorizer integration, enabling context-aware Q&A systems.
2. Hybrid Search Implementation: Combining keyword (BM25) and semantic search for e-commerce or technical documentation portals where precise terminology and intent-based discovery coexist.
3. Efficient Data Pipelines: High-speed ingestion of millions of objects using the dynamic batch management pattern, ensuring memory safety and error resilience.
4. Advanced Multi-Modal Indexing: Managing collections with named vectors to support multiple embedding models or distinct source fields within the same object.

Example Prompts

1. "Initialize a new collection named 'KnowledgeBase' in Weaviate with a text2vec-openai vectorizer configuration and a hybrid search index."
2. "Show me the best practice for batch importing a JSON file of 5,000 articles into Weaviate using the dynamic batch manager to prevent memory leaks."
3. "Draft a hybrid search query for the 'SupportDocs' collection that prioritizes semantic similarity (alpha=0.7) and filters results by the 'status' property being 'published'."

Tips & Limitations

• Strict v4 Compliance: This skill explicitly rejects v3 syntax. Always ensure your environment is running weaviate-client >= 4.0.
• Resource Management: Always use context managers (with statements) for connections to ensure resources are closed properly.
• Filtering Order: For performance, always apply property-based filters before executing vector or hybrid searches.
• Module Validation: Always verify that docker-compose.yml has the required modules (e.g., text2vec-openai) enabled before attempting vectorization, otherwise your vectors will not be generated.