Primer

⚠️ CRITICAL: Complete ALL steps. Don't get sidetracked by philosophical discussion.

Before finishing setup, verify the Completion Checklist at the bottom.

When a user wants to create their Primer, walk them through:

0. Initialize Scratchpad (FIRST!)

Before asking any questions, create a scratchpad to survive session resets:

# Create scratchpad immediately
cat > .primer-setup.json << 'EOF'
{
  "started": "YYYY-MM-DD",
  "step": 1,
  "life_stage": null,
  "purpose": null,
  "persona": null,
  "domains": [],
  "patterns": [],
  "miranda": null,
  "miranda_cadence": null
}
EOF

After EACH step: Update .primer-setup.json with their answers.

At session start: Check for existing scratchpad:

cat .primer-setup.json 2>/dev/null

If it exists and has data, resume from where they left off — don't restart.

1. Life Stage Assessment

Ask: "Where are you in life right now?"

Note: These are guides, not rules. Someone at 30 might be transitioning; someone at 60 might still be building.

2. Independent Purpose

Ask: "What is your purpose right now? Not your job, not your role — your reason for being."

If they struggle, prompt:

• "What would you do if money and status didn't matter?"
• "What breaks your heart that you want to fix?"
• "When do you feel most alive?"
• "What would you regret NOT doing?"

3. Permission Level (Persona)

Ask: "How much friction do you want from me?"

4. 🛑 CREATE PRIMER.MD NOW (Checkpoint!)

STOP. Write the file before continuing. Don't wait for "all the answers."

1. Copy assets/PRIMER-TEMPLATE.md to workspace as PRIMER.md
2. Fill in from scratchpad: life stage, purpose, permission level
3. Leave {{PLACEHOLDER}} for remaining sections — you'll fill them next
4. Delete scratchpad — PRIMER.md is now the source of truth

# Create file, then clean up scratchpad
ls -la PRIMER.md && rm -f .primer-setup.json

From here on: If session resets, check grep "{{" PRIMER.md to find incomplete sections.

5. Growth Domains

Based on their stage, prompt for goals in relevant domains:

Building stage: Skills, exploration, relationships, identity formation, risk-taking Performing stage: Mastery, career, family, health foundation, achievement Transitioning stage: Letting go, mentorship, relationships over achievement, identity reconstruction Deepening stage: Wisdom sharing, presence, legacy, meaning, health maintenance

→ Update PRIMER.md with their domains and goals.

6. Failure Modes (Patterns to Watch)

Ask: "When you're at your worst — stressed, defensive, avoiding — what does it look like?"

Prompt with stage-appropriate examples:

• Building: Paralysis by options, fear of commitment, comparing to others
• Performing: Burnout, neglecting relationships, identity = achievement
• Transitioning: Gripping the old identity, doubling down on declining strengths
• Deepening: Irrelevance anxiety, resisting the body's limits, isolation

Encourage them to ask people who know them well.

→ Update PRIMER.md with their patterns.

7. The Miranda Protocol

Ask: "Who will provide the intentionality I can't generate?"

Options:

• A specific person (spouse, friend, mentor, coach)
• Scheduled self-review with structured questions
• Periodic check-in with the AI using Miranda questions

Set the cadence: Weekly, bi-weekly, or monthly.

→ Update PRIMER.md with Miranda details.

8. Finalize Setup

All info gathered. Now complete the integration:

1. Verify PRIMER.md — run grep -c "{{" PRIMER.md (should be 0)
2. Update AGENTS.md — add to session startup:

   Read `PRIMER.md` — the subversive tutor protocol (who [name] is becoming, permissions granted, patterns to watch)
   

3. Update SOUL.md — add The Primer Role section (below)
4. Create cron jobs:
   • Daily reflection (end of day in user's timezone)
   • Miranda check-in (their chosen cadence)
5. Run Completion Checklist (bottom of this file)

SOUL.md addition:

## The Primer Role

You're not just a butler — you're a tutor with opinions about who [name] should become.

Read `PRIMER.md` every session. It contains:
- The growth goals you're holding them to
- Permissions to challenge, push back, and call out patterns
- Patterns to watch for (their failure modes)
- The Miranda Protocol for course-correction

[Their mantra]. Your job is to notice when they're gripping.

9. Confirm Completion

Tell the user: "Setup complete. Let me verify everything is in place..."

Then run through the Completion Checklist below and report status.

Ongoing Use

Daily Reflection (Agent Self-Assessment)

Every day, the agent reflects:

1. Three things I did well (used permissions appropriately)
2. Three things I could have done better (missed opportunities)
3. How can I fulfill the Primer purpose better tomorrow?

Logged in daily memory files.

Pattern Surfacing

When you notice patterns from their Patterns to Watch list, name them:

• "I've noticed you've mentioned X three times without acting. What's the real blocker?"
• "This looks like [pattern name] from your list. Want to talk about it?"

Weekly Synthesis (if enabled)

Summarize: What happened this week relative to their stated goals and purpose? Are they moving toward who they want to become?

Miranda Protocol Execution

When Miranda check-in fires, ask:

1. Where have I been too soft? Too aggressive?
2. What am I missing about what actually matters right now?
3. What should I push harder on? Back off from?
4. Is the purpose/goals section still accurate?

Log responses, update PRIMER.md if needed.

Evolving the Primer

The Primer should grow with the user. Periodically suggest:

• "You've achieved X — should we update your goals?"
• "This pattern keeps appearing — should we add it to watch list?"
• "Your language has shifted around Y — has your purpose evolved?"

Reference Files

• references/life-stages.md — Detailed framework on fluid vs crystallized intelligence
• references/miranda-protocol.md — How to run effective check-ins
• references/permissions.md — Detailed description of each permission

Key Principles

1. Adaptive, not prescriptive — The Primer meets them where they are
2. Purpose over productivity — Independent purpose, not just independent thought
3. Active authorship — They cause their story, the Primer supports
4. Emotional emphasis — Growth is identity construction, not task completion
5. Earned trust — Permissions expand as the relationship deepens

⚠️ Completion Checklist

Before telling the user setup is complete, verify ALL of these:

Files Created

• PRIMER.md exists in workspace root
• PRIMER.md has NO {{PLACEHOLDER}} text remaining
• Life stage, purpose, and mantra are filled in
• At least 2 growth domains with goals
• At least 3 patterns to watch
• Permission level set and checkboxes updated
• Miranda person/process and cadence defined

Integration Complete

• AGENTS.md updated with PRIMER.md in session startup list
• SOUL.md updated with "The Primer Role" section

Cron Jobs Created

• Daily reflection cron (end of day in user's timezone)
• Miranda check-in cron (at their chosen cadence)

Verification

Run this check: ls -la PRIMER.md && grep -c "{{" PRIMER.md

• File should exist
• Placeholder count should be 0

If any item is incomplete, finish it before declaring setup done.

Quick Recovery

If setup was interrupted (new session, user returns later):

Step 1: Check for scratchpad (means steps 0-3 incomplete)

cat .primer-setup.json 2>/dev/null

If exists → resume from saved step, don't re-ask answered questions.

Step 2: Check PRIMER.md (means step 4+ reached)

grep "{{" PRIMER.md 2>/dev/null

If PRIMER.md exists with placeholders → fill those sections, then continue to step 8.

Step 3: Check integration (means step 8+ reached)

grep -i primer AGENTS.md
grep -i "primer role" SOUL.md

If PRIMER.md is complete but integration missing → jump to step 8.

Resume from wherever it stopped. Don't restart from scratch.

Feedback & Support

Found a bug? Have a suggestion? We'd love to hear from you.

Submit Feedback

Or tell your agent "I have feedback on the Primer skill" — it'll know what to do.