What problem does it solve?
Auto-compaction triggers at arbitrary points during long AI sessions, often interrupting mid-task and losing critical context that is expensive to re-acquire, especially for multi-phase workflows like research, planning, and implementation.
Core Features & Use Cases
- Logical Compaction Timing: Suggests manual
/compact only at natural task phase boundaries instead of random auto-trigger points.
- Tool Call Threshold Tracking: Monitors tool usage to prompt compaction at configurable intervals, with periodic reminders after the initial threshold.
- Phase-Based Decision Guide: Provides a clear reference table to help users decide when compaction is appropriate based on their current workflow stage.
- Use Case: For a developer working on a multi-phase feature build, the skill will prompt them to compact after finishing research and finalizing their implementation plan, preserving the plan while clearing bulky research context to free up space for coding.
Quick Start
Use the strategic-compact skill to identify optimal compaction points for your current long-running multi-phase task to avoid losing critical context mid-work.