User Profile

USER.md tells your agent who you are. It reads this at the start of every session. Name, nickname, timezone, goals, how you work -- whatever context helps it help you.

Setup Prompt

Create a USER.md file with my info:
- Name: [your full name]
- What to call me: [nickname]
- Timezone: [e.g., America/Chicago]
- Location: [city]
- Social profiles: [Twitter, GitHub, etc.]
- Key context: [what you want help with]
- Current goals: [your top 2-3 goals]
- Notes: [anything else relevant -- work style, preferences, etc.]

What to Include

The more context you give, the better. Some things worth adding:

• Work style -- "I do deep work in the morning, meetings in the afternoon"
• Communication preferences -- "Be direct, skip the pleasantries"
• Values -- what matters to you (your agent can flag when you're drifting)
• Personality assessments -- Enneagram, MBTI, etc. These actually help the agent understand your patterns.
• Growth edges -- areas you're actively working on

Advanced: Personality Pattern Flagging

I went pretty deep on this. I fed Chewy all my personality assessment data (Enneagram, Golden Personality Profiler, values assessment, Birkman) and asked it to flag patterns when it sees them in conversation. For example, my Type 5 tendency to withdraw and conserve energy -- Chewy will call it out with something like "That sounds like scarcity thinking -- is it real?"

Add a section to MEMORY.md for pattern flagging. Based on my
personality profile, flag these patterns when you see them:
- [pattern 1] -- brief description of what to watch for
- [pattern 2]
- etc.

How to flag: Brief, non-preachy. Just name it and move on.
Example: "That sounds like scarcity thinking -- is it real?"