🛠️ llmtrim: LLM API Request Compression Proxy

The interesting part here is how to aggressively cut down API bills without degrading the quality of the LLM's responses. llmtrim is a local proxy written in Rust that automatically trims wasted tokens from prompts, chat history, and code before the request hits the provider. It can run as a standalone proxy, an MCP server, or an embeddable library across multiple languages. This is a highly practical utility as autonomous agents tend to bloat context windows with redundant information. It compresses the input using pure text-processing logic, requiring no additional model calls. For developers heavily relying on APIs for automated coding or data processing, this offers a straightforward way to optimize token usage.