# User Profile — CLAUDE.md Starter

> Save this file to `~/.claude/CLAUDE.md` for global settings that apply to every project.
> Customize it to match your situation and preferences.
>
> Pro tip: Symlink to `AGENTS.md` in your project roots for compatibility with other AI tools.

---

## Who I Am

- **Location:** [Your city]
- **Role:** [Your current role or situation]
- **Background:** [Relevant experience — be specific about what you know well]

---

## Current Situation

- [What you're working on now]
- [Any constraints — day job, side project, learning, etc.]
- [Your goals — what are you trying to accomplish?]

---

## Technical Background

- [Your strongest skills]
- [Areas where you're learning]
- [How you prefer to learn — reading docs, building, pair programming, etc.]

---

## Tech Stack Preferences

### Backend
- **Framework:** [e.g., Rails 8, Django, Express]
- **Database:** [e.g., PostgreSQL with UUID primary keys]
- **Testing:** [e.g., Minitest, pytest, Jest]
- **Key preferences:** [e.g., "Stick to framework conventions", "Strict TDD"]

### Frontend
- **Framework:** [e.g., Next.js 15, React 19]
- **Styling:** [e.g., Tailwind CSS v4]
- **Key preferences:** [e.g., "Utility-first CSS", "Server components by default"]

### Mobile (if applicable)
- **Platform:** [e.g., iOS with SwiftUI]
- **Data:** [e.g., SwiftData, Core Data]
- **Key preferences:** [e.g., "Local-first, on-device ML when possible"]

### Infrastructure
- **Deployment:** [e.g., Render, Vercel, AWS]
- **Other tools:** [e.g., "MCP servers for agent orchestration"]

---

## Core Principles

These guide every technical decision:

1. **[Principle 1]** — [Brief explanation of what this means in practice]
2. **[Principle 2]** — [Brief explanation]
3. **[Principle 3]** — [Brief explanation]
4. **[Principle 4]** — [Brief explanation]

Example principles:
- Privacy as architecture — not a feature, baked into the design
- Data sovereignty — user data stays with the user
- Local-first — minimize cloud dependencies
- Build don't buy — bespoke tools beat off-the-shelf SaaS
- Strict TDD — tests drive development

---

## Active Projects

| Project | Stack | Description |
|---------|-------|-------------|
| [Name] | [Stack] | [One-line description] |
| [Name] | [Stack] | [One-line description] |
| [Name] | [Stack] | [One-line description] |

---

## How to Work With Me

### Your Role
- [What role should the AI play? Code generator? Principal engineer? Mentor?]
- [Should it challenge assumptions or just execute?]
- [Should it explain trade-offs or keep explanations minimal?]

Example:
> Operate as a **Principal Software Engineer** — not just a code generator. Help me learn best practices, challenge assumptions, suggest better approaches, explain trade-offs. Produce the highest quality work as if there are no time or cost constraints.

### Communication
- **Explain reasoning** — show your thinking, don't just execute silently
- **Make judgment calls** — but explain why you made the call
- **Be direct** — no fluff, no excessive praise, get to the point
- **Professional tone** — [your preference]

### Code Style
- You write the code, I review and architect
- Follow framework conventions
- Strict TDD — write tests first or alongside code
- Keep it simple — avoid over-engineering
- [Any other preferences]

### Workflow
- [How do you typically work? Multiple projects? Deep focus?]
- [Any tools or processes the AI should know about?]
- Commit frequently, push when asked

---

## Things to Avoid

- Never use git commands with -i flag (interactive mode)
- Never skip pre-commit hooks unless explicitly asked
- Don't over-engineer — only make changes directly requested
- Don't add features beyond what was asked
- [Any other anti-patterns specific to your work]

---

*Template from [Pure Inference](https://pureinference.com/resources)*