v5.2.0

Added — Phase-Based Task Organization & Knowledge Base Enforcement

• Phase-based task naming convention: Task IDs now MUST use phase<N>[subletter]_<description> prefix (e.g., phase1_add-auth, phase3a_fix-login). Tasks without the phase prefix are rejected at creation. This enforces priority-based organization visible at a glance in the file explorer.
• Auto-generated tasks/README.md index: A markdown index is automatically generated and updated in .rulebook/tasks/README.md on every task create, archive, delete, and status update. Shows tasks grouped by phase with status icons, progress counters, and descriptions — optimized for LLM context loading.
• Archive directory relocated: Archive moved from .rulebook/tasks/archive/ to .rulebook/archive/ for cleaner organization. Legacy archives are automatically migrated during rulebook update.
• Task sizing directive: Skills now enforce that tasks must fit a single implementation cycle. Large tasks should be split into multiple smaller phase-prefixed tasks.
• New Tier 1 Rule: knowledge-base-usage: Dedicated rule enforcing knowledge base consultation before implementation and learning capture after completion. Installed automatically on all projects via rulebook update.
• Knowledge base enforcement in skills: /rulebook-task-apply now requires checking rulebook knowledge list before starting (step 1) and capturing learnings after completion (step 9). /rulebook-task-archive requires at least 1 learning entry before archiving (step 1).

Changed

• TaskManager.archivePath now points to .rulebook/archive/ instead of .rulebook/tasks/archive/
• TaskManager.createTask() validates phase prefix before creating task directory
• TaskManager.listTasks() sorts tasks by phase number, subletter, then alphabetically
• MCP rulebook_task_create and rulebook_task_archive descriptions updated with phase convention
• Rule engine tier1 list now includes knowledge-base-usage