---
name: "weaviate-graphql-vector-hybrid"
description: "Combine BM25 keyword search with dense vector similarity using Weaviate Hybrid Search."
metadata:
  author: "Weaviate"
  version: "1.0.0"
  runtime_compatibility:
    - "codex"
    - "claude-code"
    - "cursor"
    - "windsurf"
triggers:
  - "setup weaviate hybrid search"
  - "configure weaviate schema"
  - "tune hybrid alpha weaviate"
permissions:
  fs_read: true
  fs_write: "workspace"
  shell_exec: false
  net_access: false
---

# Weaviate Hybrid Search & Generative AI Modules

Deploys Weaviate vector search engine: configuring data schemas with inverted indexes, hybrid search (alpha parameter weighting), cross-encoders for reranking, and multi-tenancy partitions.

## When to Invoke
Invoke this skill whenever you need to:
- setup weaviate hybrid search
- configure weaviate schema
- tune hybrid alpha weaviate

## Execution SOP & Rules
1. Connect using weaviate-client v4: client = weaviate.connect_to_local().
2. Define collection properties with vectorizer_config=Configure.Vectorizer.text2vec_openai().
3. Execute hybrid query: collection.query.hybrid(query='latency benchmarks', alpha=0.5, limit=5).
4. Inspect score fusion results to ensure relevant keyword hits match semantic vector hits.

## Verification Checklist
- Run local linting, syntax validation, and type-checks before reporting completion.
- Ensure all created or modified files adhere to workspace boundaries and avoid uncommitted secrets.
- Verify tests or build commands exit cleanly with return code 0.