Using Claude Code¶

This system is designed to be used conversationally with Claude Code in your terminal.

Getting Started¶

cd life-system
claude

That's it. Claude reads the .claude/CLAUDE.md file automatically and understands the entire system.

Talking to Claude¶

Just talk naturally. You don't need special commands.

Examples:

"Let's work on my finance goal"
"Weekly review"
"I have an idea I want to explore"
"What does my health goal look like?"
"Help me structure this into a goal"

Claude will read the relevant framework docs and operate in the appropriate mode.

The Two Spaces¶

This system has two distinct spaces for working with ideas:

flowchart LR
    subgraph visionary ["VISIONARY SPACE"]
        V1["Explore freely"]
        V2["Dream without constraints"]
        V3["No pressure to act"]
    end

    subgraph pipeline ["THE PIPELINE"]
        P1["Architect<br/>Structure idea"]
        P2["Integrator<br/>Check fit"]
        P3["Decision<br/>Pursue/Defer/Decline"]
    end

    visionary -->|"Explicit<br/>confirmation"| pipeline
    P1 --> P2 --> P3

    style V1 fill:#ab47bc,stroke:#8e24aa,color:#fff
    style V2 fill:#ab47bc,stroke:#8e24aa,color:#fff
    style V3 fill:#ab47bc,stroke:#8e24aa,color:#fff
    style P1 fill:#42a5f5,stroke:#1e88e5,color:#fff
    style P2 fill:#66bb6a,stroke:#43a047,color:#fff
    style P3 fill:#ffc107,stroke:#ffb300,color:#000
    style visionary fill:#f3e5f5,stroke:#ab47bc
    style pipeline fill:#e3f2fd,stroke:#42a5f5

Visionary Space (The Think Room)¶

Purpose: Pure creative exploration. Dream, expand, explore without any pressure.

This is where you: - Explore what life could look like - Think about big visions (living in Europe, financial freedom, career changes) - Discuss possibilities without committing to anything - Let ideas breathe and evolve over multiple conversations

Key features: - Ideas can live here forever - No feasibility checks - No alignment questions - Visions can span multiple goals and domains - Claude explores the texture of ideas, not just logistics

Example conversation:

"My wife and I want to live in Europe someday. Let's explore what that actually looks like."

Claude will ask about the feel of that life, what you're reaching for, what opens up — not "how will you afford it?"

The Pipeline (Formalization)¶

Purpose: Turn ideas into concrete goals when you're ready.

This is where ideas get: - Structured into GSA format (Goal, Strategy, Actions) - Checked for fit with your life - Decided on: pursue, defer, or decline

Key feature: Entry requires explicit confirmation. Claude will always ask:

"Just to confirm — you want to move this into the pipeline? Ready to proceed?"

This prevents accidentally leaving the creative space.

Agent Modes¶

Claude operates in different modes depending on what you need.

Visionary Mode¶

Purpose: Dream and explore without constraints.

Activate with: - "Visionary mode" - "Let's explore..." - "I want to dream about..." - "What if..." - "I've been thinking about..."

What Claude does: - Asks "what if?" and "what else?" - Explores the bigger version of ideas - Asks what you're really reaching for - Explores the texture and feel, not just logistics - Never evaluates feasibility - Never checks alignment - Lets ideas breathe

Visions vs Goals: A vision isn't a goal. "Living in Europe" might spawn goals across finance, career, and relationships. In Visionary mode, we explore holistically.

Architect Mode¶

Purpose: Ground ideas into concrete structure.

Activate with: - "Architect mode" - "Structure this..." - "What would this actually take?" - "Turn this into a goal"

What Claude does: - Shapes vague ideas into GSA goals - Identifies requirements and dependencies - Estimates scope realistically - Creates outcomes and KPIs - May pull from earlier visionary exploration

Note: If triggered from visionary space, Claude will ask for confirmation first.

Integrator Mode¶

Purpose: Check if something fits with your life.

Activate with: - "Integrator mode" - "Does this fit?" - "Check capacity" - "What are the trade-offs?"

What Claude does: - Checks alignment with direction - Evaluates against DeCO filter (coupling, compounding, optionality) - Assesses capacity realistically - Surfaces conflicts with existing goals - Recommends: pursue, defer, or decline

Reviewer Mode¶

Purpose: Facilitate reviews conversationally.

Activate with: - "Weekly review" - "Monthly review" - "Quarterly review" - "Let's review my goals"

What Claude does: - Walks through each domain conversationally - Asks open-ended questions (not just "did you hit the target?") - Goes deeper where there's friction or energy - Captures insights authentically - Generates review documents

Reviews are conversations, not forms. You talk, Claude listens and captures.

Assessor Mode¶

Purpose: Analyze patterns and progress over time.

Activate with: - "Assessor mode" - "Assess my progress" - "What patterns do you see?" - "Give me an honest assessment"

What Claude does: - Reviews data across multiple review periods - Identifies patterns and trends - Flags what's working and what's not - Diagnoses potential layer violations - Makes evidence-based recommendations

Ideation Workflow¶

Starting an Exploration¶

Just start talking:

"I've been thinking about what it would be like to have completely passive income..."

Claude enters Visionary mode automatically. Explore freely.

Saving an Exploration¶

At natural pause points, Claude offers:

"Want me to capture this exploration so we can return to it?"

If yes, it's saved to docs/personal/ideation/visionary/ with: - The vision narrative - Key themes and domains it touches - Questions explored - What you're reaching for

Returning to an Exploration¶

"Let's pick up that conversation about living in Europe"

Claude finds the file and continues where you left off.

Entering the Pipeline¶

When you're ready to formalize:

"I want to actually pursue this" or "Let's make this real"

Claude confirms:

"Just to confirm — you want to move this into the pipeline? This will take us through structuring it as a goal and checking if it fits. Ready?"

After you confirm, Claude automatically: 1. Architect → Structures into GSA format, outcomes, KPIs 2. Integrator → Checks alignment, DeCO filter, capacity, trade-offs 3. Decision → Asks: "Pursue, defer, or decline?"

Based on your decision: - Pursue → Goal files created, idea archived to approved/ - Defer → Archived with trigger conditions ("revisit when X") - Decline → Archived with reasoning

Review Workflow¶

Weekly Review (20-30 min)¶

"Weekly review"

Claude walks through each domain: - How did this area feel this week? - What moved forward? What didn't? - Where was there friction? - What's the focus for next week?

At the end, Claude generates a review doc in docs/personal/reviews/weekly/.

Monthly Review (30-60 min)¶

"Monthly review"

Deeper look at outcomes: - Progress against KPIs - Deadline reality check - Patterns across weeks - Capacity alignment

Quarterly Review (60-90 min)¶

"Quarterly review"

Goal-level evaluation: - Direction still valid? - Each goal still worth its cost? - Capacity reality check - Select quarterly focus

Annual Review (Half day)¶

"Annual review"

Life design level: - Year reflection - Goal closures - Direction update - Next year theme

File Locations¶

Content	Location
Goals	`docs/personal/goals/{domain}/goal.md`
Outcomes	`docs/personal/goals/{domain}/outcomes.md`
KPIs	`docs/personal/goals/{domain}/kpis.md`
Weekly reviews	`docs/personal/reviews/weekly/YYYY-MM-DD.md`
Monthly reviews	`docs/personal/reviews/monthly/YYYY-MM.md`
Quarterly reviews	`docs/personal/reviews/quarterly/YYYY-QN.md`
Annual reviews	`docs/personal/reviews/annual/YYYY.md`
Visionary explorations	`docs/personal/ideation/visionary/`
Pipeline processing	`docs/personal/ideation/processing/`
Decided ideas	`docs/personal/ideation/archive/{approved,declined,deferred}/`
Direction	`docs/personal/direction.md`

Tips¶

You don't need to be precise. Claude figures out what mode to use based on context.
Say what you want, not how. "Help me think through my career" is better than "activate architect mode."
Visions can be big. A single vision might touch multiple domains. That's fine — explore it holistically first.
Reviews are conversations. You don't fill out forms — you talk, Claude captures.
Pipeline entry is explicit. Claude will always confirm before leaving visionary space.
Push to GitHub after changes. The site auto-updates.

Common Commands¶

What you want	What to say
Explore an idea	"I've been thinking about..." / "What if..."
Save an exploration	"Save this" / "Capture this"
Return to an exploration	"Let's pick up that conversation about X"
Turn idea into goal	"I want to pursue this" (Claude will confirm)
Continue in-progress pipeline	"Let's continue with that idea about X"
See saved ideas	"What's in my visionary space?"
See in-progress pipeline items	"What's in processing?"
Weekly review	"Weekly review"
Monthly review	"Monthly review"
See a goal	"Show me my health goal"
Update a goal	"Update my health outcomes"
Assess progress	"How am I doing on health?" / "Assess my progress"
Check all goals	"What are all my current goals?"