Weaviate Launches Agent Skills to Empower AI Coding Agents

Arina Makeeva Avatar
Illustration

On February 20, 2026, Weaviate, a frontrunner in the open-source AI database arena, unveiled an exciting new feature known as Weaviate Agent Skills. This development represents a significant step forward in providing coding agents, such as Claude Code, Cursor, GitHub Copilot, VS Code, and Gemini CLI, with specialized tools designed to generate production-ready code that targets Weaviate workflows.

This launch builds upon the foundation laid by Weaviate’s Query Agent, which was first showcased in March 2025 and achieved general availability by September 2025. The Query Agent allows for the execution of natural language queries across multiple collections, introducing features like multi-collection routing, intelligent query expansion, and the capability to manage user-defined filters. These enhancements enable developers to extract optimal results from even the most complex questions. For practical exploration, developers are encouraged to utilize Weaviate Cloud’s free Sandbox clusters, which provide an arena for experimentation with small instances that remain active for 14 days and can be upgraded to production-level Shared Cloud setups.

Comprehensive Repository Tools

The launch includes a robust repository structured on GitHub, specifically at github.com/weaviate/agent-skills. This repository is divided into two main sections, designed to support the entire lifecycle from straightforward operations to complex applications. The first part, available at /skills/weaviate, encompasses various scripts tailored for essential tasks related to cluster management, data lifecycle operations, and sophisticated retrieval techniques.

Cluster management includes functionalities such as schema inspection, collection creation, and metadata retrieval. For data operations, the repository facilitates imports of various data formats, including CSV, JSON, and JSONL files, alongside tools for generating example datasets. Advanced search capabilities leverage the Query Agent’s power, offering options for hybrid searches that blend semantic and keyword queries, ensuring users can tailor their experience effectively.

Cookbooks for Production Applications

The second section of the repository, located at /skills/weaviate-cookbooks, supplies developers with essential blueprints for creating production applications. Among the highlights are end-to-end implementations featuring Query Agent chatbots. These advanced applications utilize FastAPI for backend functionalities and Next.js for frontend interactions, making them lightweight and efficient. Additionally, the repository showcases multimodal PDF Retrieval-Augmented Generation (RAG) pipelines utilizing ModernVBERT for multivector embeddings, in conjunction with generation tools like Ollama or Qwen3-VL.

These resources lay the groundwork for various implementations, from basic to advanced systems, including agentic RAG functionalities equipped with decomposition and reranking capabilities. Developers can also leverage DSPy-optimized agents that utilize custom tools and leverage persistent memory.

Six Streamlined Slash Commands

A particularly innovative feature of Weaviate Agent Skills is the introduction of six intuitive slash commands that coding agents can auto-discover and execute, streamlining interactions with Weaviate. These commands facilitate various actions:

  • /weaviate:ask: Generates AI-driven answers with citations sourced from the Query Agent.
  • /weaviate:collections: Lists all existing schemas or inspects specific collections.
  • /weaviate:explore: Displays data metrics, counts, and sample objects.
  • /weaviate:fetch: Retrieves specific objects by ID or applies property filters.
  • /weaviate:query: Executes natural language searches across multiple collections.
  • /weaviate:search: Conducts hybrid, semantic, and keyword searches with adjustable parameters like alpha blending.

These streamlined commands empower developers to easily interact with Weaviate’s vast capabilities. For example, a user can execute a command such as “/weaviate:search query ‘best laptops’ collection ‘Products’ type ‘hybrid’ alpha ‘0.7’” to retrieve balanced results relevant to their needs.

Weaviate’s launch of Agent Skills sets a new standard for coding flexibility and efficiency in AI database management. As the integration of AI into everyday business practices accelerates, developments like these will empower developers, business leaders, and investors alike, allowing them to harness the true potential of AI-driven coding frameworks.

Leave a Reply

Your email address will not be published. Required fields are marked *