Pi Cache Optimizer

A Pi extension that improves KV cache and prompt cache hit rates for AI providers like DeepSeek, OpenAI, Claude, and Gemini. v2.5.6 (stable) maintained by jiangge. Key features: reorders prompts to put stable content first, compresses Pi skill lists, injects prompt_cache_key fallback for OpenAI-compatible requests, warns about proxy cache/compat issues, and displays session-scoped cache stats in the footer. Renamed from pi-deepseek-cache-optimizer for broader provider support. Uses peer dependency on @earendil-works/pi-coding-agent. Does not modify models.json. Opt-out via environment variables for prompt rewriting, skill compression, or OpenAI cache key fallback.