技能 provider-management
🔄

provider-management

安全 🌐 網路存取⚙️ 外部命令📁 檔案系統存取

Switch AI providers automatically with smart fallback

Managing multiple AI providers with different authentication methods and usage limits creates reliability challenges. This skill provides centralized provider management with automatic fallback based on authentication priority and available capacity.

支援: Claude Codex Code(CC)
📊 71 充足
1

下載技能 ZIP

2

在 Claude 中上傳

前往 設定 → 功能 → 技能 → 上傳技能

3

開啟並開始使用

測試它

正在使用「provider-management」。 Show me the status of all configured AI providers

預期結果:

  • Anthropic providers: subscription (100), OAuth (50), API (10) - all configured
  • OpenAI providers: API (10) configured, OAuth not set up
  • Google providers: API (10) configured, OAuth not set up
  • Daily usage this month: 2.3M tokens total
  • Auto-switch enabled: yes

正在使用「provider-management」。 Configure provider priority for Claude models

預期結果:

  • Provider priority order set for Claude models:
  • 1. Anthropic subscription (priority 100)
  • 2. Anthropic OAuth (priority 50)
  • 3. Anthropic API (priority 10)
  • 4. Bedrock (priority 10)
  • Auto-fallback enabled: true

正在使用「provider-management」。 Set usage limits for my OpenAI API

預期結果:

  • Usage limits configured for openai-api:
  • Daily limit: 500,000 tokens
  • Monthly limit: 5,000,000 tokens
  • Current daily usage: 125,000 tokens
  • Warning threshold: 80%

品質評分

38
架構
100
可維護性
85
內容
30
社群
100
安全
91
規範符合性

你能建構什麼

Multi-vendor AI infrastructure

Configure multiple authentication methods per vendor to ensure continuous availability across different AI services

Cost optimization

Set usage limits and priorities to control costs while maintaining access to multiple AI models

Personal AI toolkit

Manage subscription, OAuth, and API key access across all major AI providers from one interface

試試這些提示

Check provider status 
Show me the status of all configured AI providers including their authentication type and usage
Add provider authentication 
Set up Anthropic API authentication with my API key starting with sk-ant-
Configure priority order 
Set my provider priority to use Anthropic subscription first, then Anthropic OAuth, then Anthropic API, then Bedrock
Set usage limits 
Set a daily limit of 1 million tokens and monthly limit of 10 million tokens for my Anthropic API provider

最佳實務

  • Configure at least two authentication methods per vendor for reliability
  • Set usage limits conservatively at 80% of your actual quota
  • Prioritize subscription authentication when available for best features

避免

  • Using only one authentication method creates a single point of failure
  • Ignoring usage limits can lead to unexpected billing
  • Disabling auto-switch reduces reliability when primary providers hit limits

常見問題

Which AI providers are supported?
All major providers including Anthropic, OpenAI, Google, xAI, Meta Llama, DeepSeek, Mistral, Cohere, and Qwen through services like Bedrock, Vertex, and OpenRouter.
What authentication methods work?
Three methods: subscription tokens for Pro/Plus plans, OAuth 2.0 for personal accounts, and direct API keys for programmatic access.
How does automatic fallback work?
When the primary provider reaches its usage limit, the system automatically tries the next provider in priority order until a working one is found.
Is my credential data safe?
Credentials are stored locally with restricted file permissions (0600). This skill only manages configuration - actual credential storage depends on the host application.
Why is the wrong auth method being used?
The system always prefers subscription over OAuth over API. Check your priority settings with provider-priority command and ensure limits are not exhausted.
How is this different from using providers directly?
This skill provides unified management across all providers. You configure once and the system handles switching, limits, and fallback automatically.

開發者詳情

授權

MIT

儲存庫

https://github.com/AttendantLion8/provider-fallback-plugin/tree/master/skills/provider-management

引用

master

檔案結構

📄 SKILL.md