# Manage AI Provider Fallbacks

Teams using many AI providers need predictable routing, limits, and authentication choices. This skill documents provider priority, model mappings, credential setup, usage caps, and fallback behavior.

## Install

```bash
npx skillstore add attendantlion8/provider-management
```

## Metadata

- - Slug: attendantlion8-provider-management
- - Version: 2.0.0
- - Author: AttendantLion8
- - GitHub username: AttendantLion8
- - License: MIT
- - Repository: https://github.com/AttendantLion8/provider-fallback-plugin/tree/master/skills/provider-management
- - Ref: master
- - Supported tools: Claude, Codex, Claude Code
- - Risk level: medium
- - Risk factors: external\_commands, filesystem
- - Quality score: 50
- - Quality tier: warning
- - Public page: https://skillstore.pages.dev/skills/attendantlion8-provider-management
- - Manifest: https://skillstore.pages.dev/api/skills/attendantlion8-provider-management/manifest

## Capabilities

- Lists supported model families and provider mappings for Claude, GPT, Gemini, Grok, Llama, DeepSeek, Mistral, Cohere, and Qwen.
- Documents authentication setup for subscription tokens, OAuth credentials, and direct API keys.
- Explains provider priority ordering by authentication method and manual provider ordering commands.
- Describes usage limit configuration and provider status checks.
- Shows automatic and manual provider switching workflows for fallback routing.

## Use Cases

- Standardize Provider Routing: Define which providers should handle each model family and how fallback should occur when a preferred route is unavailable.
- Control Usage Across Vendors: Set daily and monthly usage limits so teams can reduce surprise spend and preserve capacity for priority workflows.
- Configure Personal Auth Options: Compare subscription, OAuth, and API key setup paths for common AI providers before choosing a workflow.

## Prompt Templates

### Check Provider Status

```
Show my configured provider status and identify which providers are ready to use.
```

### Plan Provider Priority

```
Help me set a provider priority order for Claude models using subscription first, API second, and Bedrock as fallback.
```

### Set Usage Limits

```
Help me define conservative daily and monthly limits for OpenAI, Anthropic, and Google providers.
```

### Audit Fallback Strategy

```
Review my multi-provider fallback plan for reliability, cost control, credential risk, and provider-specific model mapping issues.
```

## Limitations

- The shard contains documentation only; no command implementation is included for review.
- Security of credential storage cannot be verified from the available file.
- Provider availability, model names, and subscription behavior may change over time.
- Users must supply valid provider credentials and follow each provider policy.

## Best Practices

- Use secret prompts or a trusted credential store instead of placing keys in visible command text.
- Configure at least two providers for important model families before enabling automatic fallback.
- Set usage limits below actual quotas so fallback can happen before hard provider failures.

## Anti Patterns

- Do not paste real API keys or session tokens into shared chats, tickets, or screenshots.
- Do not rely on one provider route for critical workflows when fallback is available.
- Do not assume documented model IDs remain current without checking provider release notes.

## Security Audit

- - Safe to publish: true
- - Audited at: 2026-06-28T12:25:13.19\+00:00
- - Summary: Static analysis reported many external command hits, but review shows these are Markdown inline code spans and fenced examples rather than executable scripts. The real risks are semantic: the skill documents command-line handling of API keys and session tokens, and it stores credentials and OAuth tokens under a hidden home-directory path without verifiable encryption.

## Stats

- - Views: 245
- - Downloads: 5
- - Favorites: 0
- - Popularity score: 0