strategic-compact

To prevent data loss during AI context compaction, you must strategically compact information at logical breakpoints by saving critical decisions and updating task lists before executing the compact command.

You should manually compact AI context at logical breakpoints during complex tasks rather than relying on automatic compaction, ensuring intermediate findings and critical context are preserved before the context window is cleared.

Compacting an AI context window strategically summarizes the current situation and next steps while discarding less relevant historical data, though clear guidelines help identify what specific information is lost versus what remains.

Prepare a pre-compaction checklist for a complex coding task by summarizing the current situation, saving critical intermediate findings and decisions, and updating task lists to ensure the AI can resume efficiently after context is compacted.

Strategic context management works effectively for coding tasks spanning multiple days by ensuring important intermediate findings and decisions are saved before compaction, allowing the AI to resume operations efficiently without data loss.

Relying solely on automatic AI context compaction risks losing critical intermediate findings and decisions because it lacks the pre-compaction actions required to strategically summarize the situation and update task lists at logical breakpoints.