data-provider-integration-contract

Name: data-provider-integration-contract
Availability: InStock
Author: elchananteitzwork-a11y

Enforces shared caching, error, confidence, and provenance patterns for external data-provider integrations.

Updated Jun 11, 2026

Search Tags:#caching #error handling #provenance #api integration #code standards #data provider #confidence tiers

Ready to run

npx skills add https://github.com/elchananteitzwork-a11y/supplement-intelligence --skill data-provider-integration-contract-elchananteitzwork-a11y

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: data-provider-integration-contract
Source: https://github.com/elchananteitzwork-a11y/supplement-intelligence/tree/main/.claude/skills/data-provider-integration-contract
Command: npx skills add https://github.com/elchananteitzwork-a11y/supplement-intelligence --skill data-provider-integration-contract-elchananteitzwork-a11y

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Teams integrating multiple external data providers often end up with inconsistent client code: ad hoc caches, raw unclassified errors, and data with no confidence tier or provenance trail. This Skill ensures every new or modified provider integration reuses the repository's existing provider-cache, provider-errors, confidence-mapping, and provenance patterns instead of inventing a new client pattern per provider. ## Core Features & Use Cases - Standardized Integration Checklist: Verifies that new provider clients route through the shared provider-cache layer, throw ProviderError subtypes from provider-errors.ts, and map provider data-quality signals to CONFIRMED/INDICATED/LIMITED confidence tiers. - Provenance Enforcement: Requires recording source, fetch timestamp, and raw-response reference through lib/provenance.ts so every claim is traceable. - Fallback Definition: Mandates defined behavior for provider unavailability or rate limiting (LIMITED tier, not a crash). - Use Case: When adding a new Keepa or openFDA integration, apply this contract to confirm the client wraps fetches in the existing cache, classifies errors through the shared taxonomy, and returns confidence-tiered, provenance-tracked data. ## Quick Start Ask the AI to review a new external data-provider integration module against the data provider integration contract before merging it.

Frequently Asked Questions about data-provider-integration-contract

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

FAQPage Schema

How do I integrate a new external data provider into an existing codebase?▼

Wrap the provider's fetch calls in the existing provider-cache layer, throw ProviderError subtypes from provider-errors.ts, and return a confidence tier plus provenance record with the data. Do not create a bespoke cache or let raw SDK errors reach callers.

What patterns should a data provider client implement?▼

A provider client should implement shared caching, a classified error taxonomy, confidence-tier mapping (CONFIRMED/INDICATED/LIMITED), provenance recording with source and fetch timestamp, and defined fallback behavior for rate limits or outages.

When should I not apply this integration contract?▼

Skip it for internal-only scoring math with no external call, tests that only mock a provider, and UI-only changes. For Anthropic LLM rate or credit handling, use the llm-cost-rate-governance skill instead.

How should provider rate limits and outages be handled?▼

Define explicit fallback behavior that returns a LIMITED confidence tier rather than crashing when the provider is unavailable or rate-limited. Raw fetch or SDK errors must never reach the caller unclassified.

Does this contract decide whether a data source is worth using?▼

No. It covers integration shape only. Whether a source is worth building is the research-evidence-agent's live-feasibility judgment, and claim framing belongs to confidence-tiered-extraction.