Session Reflection

Wrap up a practice session by identifying patterns and logging insights.

Core Principles

• •Identify friction — Where did AI need help?
• •Find patterns — What's working, what isn't?
• •Keep it brief — 5 minutes max
• •Track progress — Show improvement over time

Flow

Step 1: Ask what happened

Ask the user:

1. •"What did you work on during the build session?"
2. •"Did you have to intervene at any point? What caused it?"

Keep it conversational. Don't overwhelm with questions.

Step 2: Identify patterns

Based on their answers, reflect back:

• •Friction points: Where did AI slow down or need help?
• •Wins: What worked smoothly?
• •Architecture gaps: Could better structure have prevented issues?

Step 3: Quick reflection questions

Ask (user can answer briefly or skip):

1. •"What's one thing that would have made today smoother?"
2. •"Did complexity creep in anywhere?"
3. •"What do you want to try tomorrow?"

Step 4: Log to journal

Append to /Users/mikaelweiss/code/dotfiles/learning/journal.md:

## [DATE]

### Focus
[What they worked on]

### Interventions
[Times they had to step in, or "None"]

### Friction
- [Issues encountered]

### Wins
- [What worked well]

### Tomorrow
- [What to try next]

---

Step 5: Show progress

Read /Users/mikaelweiss/code/dotfiles/learning/challenges/log.md and summarize:

• •Total challenges completed
• •Average score (if enough data)
• •Trend: improving, stable, or declining
• •Strongest area (highest average)
• •Area to focus on (lowest average)

If not enough data yet, just note how many challenges completed.

Step 6: Close out

End with a brief encouragement to continue tomorrow. Keep it genuine, not sycophantic.

Anti-Patterns

• •Do NOT drag out the reflection beyond 5 minutes
• •Do NOT ask too many questions at once
• •Do NOT over-analyze — keep it practical
• •Do NOT skip logging — the journal is how progress is tracked