strategic-compact

To prevent AI from losing context during automatic compaction, trigger manual context compaction at logical task phase transitions to preserve crucial information before the context window auto-manages itself.

You should manually compact the context window at logical breaks between task phases, such as transitioning from a detailed research phase to a planning phase, ensuring research context is preserved before planning begins.

Strategic context compaction for AI productivity is the process of suggesting manual context compaction at defined phase transitions rather than arbitrary auto-compaction, requiring user intervention to execute a manual compact command.

Yes, automatic context management can discard important information in long AI sessions. Suggesting manual context compaction at logical intervals avoids losing critical information that auto-compaction might arbitrarily discard.

To preserve research context before starting the implementation phase, use this skill to prompt a manual compact command, ensuring the research context is safely preserved before the planning and implementation begins.