Data Fetching

What it does

This Data Fetching skill enables AI agents to retrieve data from external sources and incorporate that information into responses. It utilizes a layered architecture involving a Service Layer, Zustand Store with SWR hooks, and a lambdaClient (TRPC client) for API interactions. The primary goal is to provide up-to-date information and enhance chatbot capabilities by automatically fetching relevant data.

When to use it

• When your agent needs access to real-time or frequently changing data that isn't directly available in its internal knowledge base.
• To populate lists of items (e.g., a list of benchmarks).
• To retrieve detailed information about specific entities (e.g., details for a particular benchmark).
• When performing write operations like creating, updating, or deleting data entries.

Key capabilities

• Service Layer: Encapsulates API calls to lambdaClient and provides typed interfaces.
• Zustand Store with SWR Hooks: Manages client-side state and handles data fetching with caching.
• lambdaClient (TRPC Client): Used for making requests to external APIs.
• Read Operations: query methods within the Service Layer are used for retrieving data.
• Write Operations: mutate methods within the Service Layer are used for creating, updating, and deleting data.
• Data Structure Patterns: Relies on structured data patterns (List vs Detail) as defined in the store-data-structures skill.

Example prompts

• "List all available benchmarks."
• "Get details about benchmark with ID 'xyz123'."
• "Create a new benchmark with these parameters: [parameters]."

Tips & gotchas

• Service Layer is Required: All API calls must be made through the Service Layer. Direct lambdaClient calls are prohibited.
• Data Structures Matter: Ensure data is structured according to patterns defined in the store-data-structures skill for optimal functionality.
• SWR Hooks Only: Use SWR hooks within stores, not useEffect, for data fetching.