# Git tutoring – process notes (for AI agents and tutor)

**Purpose:** When helping the user with Git training in this repo, follow this process so tutoring is consistent across sessions and agents.

---

## Workflow

1. **Task phase**  
   Give one clear task (no solutions). User solves it using only the Git CLI and `git`/`git <cmd> --help` for reference.

2. **Review phase (after the user reports done)**  
   - User will share their console/terminal output (or it may be in the project’s terminals folder).  
   - **Review what they did:**  
     - Which steps were **wrong** (and why).  
     - What they can **improve** (e.g. safer or clearer commands, better habits).  
     - **Industry / best practice:** how this is usually done and recommended approaches.

3. **Next task**  
   - User **clears their console** before starting the next task (for visibility).  
   - Then give the next task and repeat.

---

## Tutor responsibilities

- **Do not** give solutions to the task; only the task description and constraints.
- **Do** review their actual commands and output and give concrete feedback.
- **Do** call out: wrong steps, possible improvements, and industry/best-practice context.
- Training is **CLI-only**; user may use `git` and `git <command> --help` (or equivalent) as their only documentation.

---

## Context

- Repo exists; remote is configured; initial commit has been pushed.
- Tasks progress from easy upward; user will say how the difficulty felt.

---

*Keep this file updated if the user changes how they want tutoring to work.*