Prompt Caching Strategy Designer

Design an optimal prompt caching strategy for our LLM application.

## Request Patterns
{{request_patterns}}

## Prompt Structure
{{prompt_structure}}

## Cost Breakdown
{{cost_breakdown}}

Analyze caching opportunities:

**Prefix Caching**
- Identify cacheable prefixes
- Calculate potential savings
- Implementation requirements

**Response Caching**
- Exact match opportunities
- Semantic similarity thresholds
- Invalidation strategy

**KV Cache Reuse**
- Multi-turn optimization
- Session management
- Memory trade-offs

**Provider-Specific Options**
- Anthropic prompt caching
- OpenAI cached endpoints
- Custom infrastructure

Provide:
- Caching architecture
- Expected savings calculation
- Implementation priority