llm-provider-contract

Community

Standardize safe, observable LLM provider calls

Authorzxc1a1a1
Version1.0.0
Installs0

System Documentation

What problem does it solve?

AgentHub needs a consistent, safe way to integrate multiple LLM providers and models without leaking provider-specific formats, secrets, or inconsistent streaming/structured-output behaviors into business code.

Core Features & Use Cases

  • Unified Provider Adapter Contract: Enforces a common LLMRequest, LLMResponse, LLMStreamEvent, and SafeLLMError shape so the rest of the system never depends on provider SDK specifics.
  • Streaming Normalization & Structured Output Guardrails: Normalizes provider streams into AgentHub events and requires local JSON Schema validation before structured data is accepted.
  • Reliability, Fallback, and Safety Policies: Defines timeout/retry/rate-limit boundaries, capability-aware fallback rules, and strict secret/config handling with error normalization and tracing requirements.

Quick Start

Instruct an adapter implementation to follow the llm-provider-contract YAML-less policy by routing every LLM call through the unified Provider Adapter, emitting only normalized LLMStreamEvent/LLMResponse objects, validating structured output with local JSON Schema, and never exposing API keys or raw provider responses.

Dependency Matrix

Required Modules

None required

Components

references

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: llm-provider-contract
Download link: https://github.com/zxc1a1a1/Multi_Agent-AgentHub/archive/main.zip#llm-provider-contract

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.