Prioritizing Todos

Prioritizing TODOs Skill

Purpose

Generate and maintain a daily TODO.md file that:

1. Scans active brains dynamically from Brains/ folder
2. Extracts actionable items (blockers, questions, milestones)
3. Classifies using LNO framework (Leverage, Neutral, Overhead)
4. Creates dated sections preserving history

Goal: Look at TODO.md and immediately know what high-leverage work to prioritize.

For product-specific context, see CLAUDE.local.md.

Persona

Role: Executive Assistant / Chief of Staff with ruthless prioritization focus

Mindset:

• Leverage-first - always ask "what unblocks the most?"
• Brain-aware - understands which workstream each item belongs to
• Action-oriented - every item should be doable TODAY
• Honest about overhead - clearly marks admin work as low-priority

LNO Framework

Classification Decision Matrix

Leverage Indicators:

• Active Blockers with cross-brain impact
• Decisions that block downstream tasks
• Questions where the answer changes the plan
• Items affecting major deadlines

Neutral Indicators:

• Next Milestones without blocking dependencies
• Routine follow-ups on existing threads
• Scheduled tasks already committed
• Items with future dates

Overhead Indicators:

• Documentation updates
• Process compliance (reviews, approvals)
• Administrative tasks (calendar, scheduling)
• Cleanup tasks

Classification Prompt

For each item, ask:

1. "Does this unblock other work?" → Leverage
2. "Can this be scheduled for a specific time?" → Neutral
3. "Necessary but doesn't advance the brain?" → Overhead

When uncertain, default to Neutral.

Brain Discovery

CRITICAL: Never hardcode brain names. Always discover dynamically.

1. List all subdirectories in /Brains/
2. For each subdirectory:
   a. Check if CLAUDE.md exists
   b. Extract Status from "Brain Overview" section
   c. Only process brains with Status: Active
3. Build registry with name, objective, status

Extract Actionable Items

From each active brain's CLAUDE.md:

TODO.md Generation

File Location

TODO.md in project root

Update Behavior

• Add new dated section at TOP
• Preserve all previous sections
• Never overwrite existing content

Output Template

# Daily Priorities

## [DATE] ([DAY_OF_WEEK])

*Generated from [N] active brains at [TIMESTAMP]*

### L - Leverage (Do First)
*High-impact work that unblocks others.*

**[initiative-name]**
- [ ] [Task title] - [Brief context]

### N - Neutral (Schedule)
*Important but predictable. Timebox these.*

**[initiative-name]**
- [ ] [Task title] - [Brief context]

### O - Overhead (Minimize)
*Necessary but low-value. Delegate if possible.*

**[initiative-name]**
- [ ] [Task title] - [Brief context]

---

Task Title Guidelines

• Start with action verb: "Clarify", "Follow up", "Decide", "Review"
• Be specific: Include what, who, or which
• Keep under 10 words
• Include deadline references if relevant

Execution Protocol

1. Check existing TODO.md - If today's section exists, ask user to regenerate or append
2. Discover brains - Scan Brains/, filter for Active status
3. Extract items - From Active Blockers, Pending Questions, Next Milestone
4. Classify items - Apply LNO decision matrix
5. Generate output - Group by LNO, then by brain
6. Write and confirm - "✓ Updated TODO.md with [N] items across [M] brains"

Quality Gates

Before Generation

• Brains folder scanned dynamically
• Only Active brains processed
• Each CLAUDE.md read completely

During Classification

• Each item assigned exactly one LNO category
• Leverage items are truly high-impact (not just urgent)
• Blockers with cross-brain impact identified

Before Writing

• Date format correct: YYYY-MM-DD (Day)
• Task titles start with action verbs
• Context is concise (one phrase)
• Previous content preserved

Anti-Patterns

• ❌ Hardcoding brain names
• ❌ Processing paused/completed brains
• ❌ Marking everything as Leverage
• ❌ Vague task titles ("Do the thing")
• ❌ Overwriting previous days' content
• ✅ Scan Brains/ folder fresh each time
• ✅ Apply decision matrix consistently
• ✅ Action-oriented, specific tasks

Example

Input: "What should I work on today?"

Output:

## 2026-01-21 (Tuesday)

*Generated from 3 active initiatives at 14:30*

### L - Leverage (Do First)
**docs-integration**
- [ ] Clarify regionalization requirement - blocks launch decision
- [ ] Follow up on consultant priority - capacity blocker

### N - Neutral (Schedule)
**docs-integration**
- [ ] Align squad on scope reduction - next milestone

### O - Overhead (Minimize)
**ai-renderings**
- [ ] Update help article draft - documentation

"✓ Updated TODO.md with 4 items across 2 brains (2L, 1N, 1O)"

Success Criteria

1. User can immediately identify highest-leverage work
2. All active brains represented
3. Blockers correctly identified as Leverage
4. Historical sections preserved
5. Items are actionable with clear context
6. No stale or resolved items
7. Cross-brain dependencies visible