compound

To embed session learnings into system conventions, you can run a dedicated skill that analyzes your session work for non-obvious facts, failure modes, and useful techniques, then automatically updates your patterns to create accumulated advantage.

Analyzing session work for failure modes and convention friction involves reviewing your recent work to capture effective techniques and identify areas where existing patterns caused resistance, allowing you to generate tasks for unresolved questions.

Yes, you need Git and Bash to audit fleet worker sessions, as Git is required for diffing changes and Bash is necessary for analyzing logs to detect gravity candidates and identify quality improvement areas.

You can generate tasks for unresolved questions by analyzing your session experiments to identify multi-phase gaps, failed criteria, and research questions, then embedding these findings as actionable tasks within your system.

You can update system patterns based on practical session experience by analyzing the work completed and embedding the discovered effective techniques directly into your skills and conventions to prevent repetitive mistakes.