pi-prefix-cache-compaction
esm
Pi compaction that reuses the server's prefix cache (vLLM / SGLang / llama.cpp Anthropic endpoints): no cold re-prefill of the history, and a warm-up so the next turn starts from cache.
Version 0.2.0 License MIT
Keywords
pi-packagepipi-extensionpi-coding-agentcoding-agentextensioncompactioncontext-compactionauto-compactprefix-cacheprompt-cachekv-cachekv-cache-reuseanthropic-messagesvllm
INSTALL