What problem does it solve?
This Skill eliminates the risk of losing critical context during long, multi-phase AI workflows caused by arbitrary auto-compaction, which often interrupts tasks mid-execution and discards valuable intermediate context.
Core Features & Use Cases
- Intelligent Compaction Timing Suggestions: Tracks tool call counts to identify natural phase transition points (such as after completing research, finalizing a plan, or finishing a milestone) and prompts you to run manual
/compact only at these logical boundaries.
- Configurable Thresholds: Lets you adjust the tool call count threshold for initial compaction suggestions and the interval for subsequent reminders to match your workflow pace.
- Decision Guidance: Includes a reference table to help you evaluate when compaction is safe (e.g., after switching from debugging to a new feature) and when it would risk losing critical context (e.g., mid-implementation).
- Use Case: If you are working on a multi-step coding project that involves research, planning, implementation and testing, this Skill will prompt you to compact your context right after you finalize your implementation plan, so you have full context for the coding phase without carrying over bulky research notes.
Quick Start
Use the strategic-compact skill to receive timely prompts for manual /compact at logical workflow boundaries during long, multi-phase tasks to preserve relevant context and avoid mid-task context loss.
