Token Saver Context Compression Alias for . Provides context window optimization by compressing large payloads before reasoning. When to Use - Context approaching budget limits (80K+ tokens) - Large file reads that can be summarized - Multi-agent pipelines with context handoff - Any scenario where would be used Activation Activate this skill when appears or when the active context is approaching the 80K token pressure threshold. Usage This skill exists as a named alias so agents referencing in their skills arrays resolve correctly. The implementation lives in . ---