Hierarchical Task Numbering System

[Vamshi-Bantaram]

When LLMs add new tasks to tasks.md, they must renumber all subsequent tasks in the entire file. This is inefficient:

• High token consumption: Adding 1 task at position 10 in a 50-task file requires renumbering 40+ tasks (2000+ tokens)
• Slow and error-prone: Sequential numbering (T001, T002, T003...) forces cascading renumbering
• Maintenance overhead: Updates require changing references throughout the document

Proposed Solution

Use hierarchical task numbering (e.g., T1.1, T1.2, T1.3) grouped by phase/story. Adding a task only requires renumbering within the same parent group.

Format Example

Current (Sequential):

- [ ] T001 Install uv
- [ ] T002 Initialize project
- [ ] T003 Add pytest
- [ ] T004 Create tests directory
- [ ] T005 Create schema parser  ← Adding here requires T005-T043 renumbering

Proposed (Hierarchical):

## Phase 1: Setup
- [ ] T1.1 Install uv
- [ ] T1.2 Initialize project
- [ ] T1.3 Add pytest
- [ ] T1.4 Create tests directory

## Phase 2: Foundational
- [ ] T2.1 Create schema parser  ← Adding T2.1.1 only affects T2.1.x tasks

Adding a new task:

## Phase 1: Setup
- [ ] T1.1 Install uv
- [ ] T1.2 Initialize project
- [ ] T1.2.1 [NEW] Verify Python version  ← Only T1.3+ need renumbering
- [ ] T1.3 Add pytest  ← Renumbered from T1.3