Explore 3 GitHub repositories focused on context-compression. Discover top-starred projects and those trending this week.
14-stage Fusion Pipeline for LLM token compression — reversible compression, AST-aware code analysis, intelligent content routing. Zero LLM inference cost. MIT licensed.
97% token reduction for AI coding sessions — zero deps, 31 languages, MCP server
A drop-in proxy that compresses bloated code context in real-time, cutting LLM API costs by 50–80% without losing what the model actually needs to know.
14-stage Fusion Pipeline for LLM token compression — reversible compression, AST-aware code analysis, intelligent content routing. Zero LLM inference cost. MIT licensed.
97% token reduction for AI coding sessions — zero deps, 31 languages, MCP server
A drop-in proxy that compresses bloated code context in real-time, cutting LLM API costs by 50–80% without losing what the model actually needs to know.