llm-db

Catalog and normalize provider and model metadata for deterministic LLM selection.

7|Updated Mar 8, 2026
One-click install
npx skills add https://github.com/arrowcircle/jido-skills --skill llm-db
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: llm-db
Source: https://github.com/arrowcircle/jido-skills/tree/main/skills/llm-db
Command: npx skills add https://github.com/arrowcircle/jido-skills --skill llm-db

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Centralizes and normalizes provider and model metadata to enable deterministic LLM selection across Elixir apps.

Core Features & Use Cases

  • Build and maintain a catalog of providers, models, capabilities, and pricing to support decision-making.
  • Normalize provider identifiers and map them to stable IDs for cross-provider routing and caching.
  • Create runnable examples and guardrails that demonstrate how catalog data feeds later req_llm or jido workflows.

Quick Start

Provide a runnable catalog lookup example that selects a model from llm_db based on provider, capability, and price constraints.

Frequently Asked Questions about llm-db

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I build an LLM model catalog in Elixir for deterministic model selection?

To build an LLM model catalog in Elixir, you centralize and normalize provider and model metadata using a stable catalog schema with explicit provider IDs, enabling deterministic selection by capability or price across Elixir apps.

How do I normalize provider identifiers for cross-provider LLM routing?

You normalize provider identifiers by mapping raw provider metadata to stable IDs within a catalog schema, which ensures consistent cross-provider routing and caching for LLM workflows in Elixir.

Can I select an LLM model by capability and price constraints in Elixir?

Yes, you can select an LLM model by capability and price constraints by querying a normalized catalog of provider metadata, applying routing logic to filter models that match your specific application requirements.

Does llm-db work with req_llm and jido workflows?

Yes, llm-db integrates with req_llm and jido workflows by providing normalized catalog data that feeds downstream routing decisions, with examples demonstrating how lookup logic translates into actionable workflow steps.

What is the best way to structure provider metadata for LLM routing decisions?

The best way to structure provider metadata for LLM routing is to implement a stable catalog schema with explicit provider IDs, normalizing capabilities and pricing to drive deterministic lookup and routing logic across providers.

When do I need a centralized model catalog instead of hardcoding LLM provider details?

You need a centralized model catalog when managing multiple providers and require deterministic selection by capability or price, replacing hardcoded provider details with normalized metadata for consistent cross-provider routing and caching.