
Creates structured planning documents for projects and initiatives, breaking down complex work into stages with clear actions and validation points.
Interactive inbox triage and email management - get to inbox zero efficiently
Re-prioritize tasks for the week - ruthless filtering to ~10
Use when user asks about their emails, tasks, calendar, project status, memory, or wants the AI Chief of Staff to handle a request
Expert consultation mode - ask meaningful questions until enough information to maximize success, then execute.
Flip core assumptions to reveal hidden constraints and alternative approaches - 'what if the opposite were true?'
Interactive brainstorming and advisory mode that helps think through questions, critique approaches, propose alternatives, and reach optimal decisions.
Facilitated challenge mode to stress-test ideas via steelman-first critique, pre-mortems, inversion, and red-teaming.
Lists all available commands, skills, subagents, hooks, and built-in keywords
Draft follow-up emails and action items after meetings.
Maintain two-tier memory system (Chief-of-Staff + topics) after conversations, extracting and organizing new facts based on utility thresholds.
Spot patterns appearing in 3+ domains to find universal principles.
Harsh critique mode - list top 3 logical fallacies or weaknesses in thinking without being polite.
Summarize patterns from recent journal entries
Explain concepts at three levels: 5-year-old, real-world analogy, and university-level definition.
Find one insight that eliminates multiple components - 'if this is true, we don't need X, Y, or Z'
Show overdue and slipping tasks that need attention
Meta-skill for creating new skills in the proper format with [BRACKET] sections and correct structure.
Focused code refactoring with senior engineering judgment. Preserves behavior, improves quality.
Scan repository for accidentally committed secrets and credentials using gitleaks.
Prepares briefing notes for meetings by detecting type (internal vs external) and gathering relevant context.
Execute a task autonomously - Claude handles it end-to-end
Draft an email for review and approval before sending
Process items in the Inbox folder - preview, suggest destinations, and file with minimal frontmatter
Create or update today's journal entry
Guided interview that populates Chief-of-Staff.md with your context — priorities, projects, clients, preferences, and personal details.
Generate a printable Markdown view of TickTick tasks for the next 7 days
Dispatch to the right problem-solving technique based on how you're stuck.
Simulate a debate between a pragmatic operator and radical futurist to find the best path forward.
Test at extremes (1000x bigger/smaller) to expose fundamental truths hidden at normal scales.
Sync state with external systems (also runs as part of /briefing)
Batch changes into logical commits with proper messages and safe concurrent handling.
Session-end self-review. Scans for corrections, failures, and missing context, then proposes concrete rule updates.
Update and open the Daily Briefing - 30 second scan of today's priorities
Force unrelated concepts together to discover emergent properties - 'What if we treated X like Y?'
Review old tasks to archive, delegate, or clean up
Surface blind spots and areas of uncertainty in thinking.
# Book Summarizer Skill Transforms a `.md` book into a rich, referenceable knowledge artifact. ## When to Use - User provides a markdown file of a book (converted from epub/pdf) - User says "summarize this book" or "extract insights from..." - Processing books for the Knowledge/Books/ folder ## Architecture **Pass 1 (Sonnet 1M):** Structure & Theme Extraction **Pass 2 (Sonnet 1M):** Quote Mining & Evidence Collection **Pass 3 (Opus):** Synthesis, Connections & Actionable Insights ## Execut
Quick status check on any project
Comprehensive web research that launches multiple search queries, synthesizes findings, and delivers detailed reports with attribution.
End-of-week reflection and planning
Guide for gathering information by asking 1-2 questions at a time with proposed answers, building progressively to avoid cognitive overload.