plans

Manage hierarchical planning documents by using token-budgeted lazy loading to load only relevant sections into AI context. This approach optimizes large markdown files by splitting oversized plans into smaller, manageable chunks.

Split oversized markdown plans using the rebalance feature, which automatically breaks down large documents with YAML frontmatter. This process divides complex project dependencies into smaller, context-ready sections.

Token-budgeted lazy loading is a mechanism that loads specific markdown sections into AI context based on token limits. It uses YAML frontmatter and automatic rebalancing to efficiently inject cascading context.

Yes, you need Python installed because this Skill relies on Python scripting to execute token counting and document rebalancing operations. It also uses markdown for documentation and YAML frontmatter for metadata.

Initialize and analyze a hierarchical plan by running the plan_manager.sh script with the init command, then use the analyze command to evaluate the document structure. Use context to load sections and update-summary to refresh metadata.

Use automatic rebalancing when your hierarchical planning documents grow too large for AI context windows. It prevents token limit overflows by splitting oversized plans and updating summaries to maintain structural integrity.