gemini-memory-sync

# Gemini Memory Sync

## Documentation Delegation

> **Documentation Source:** For authoritative memory/import syntax and current features, query `gemini-cli-docs` skill.
> This skill provides sync patterns; `gemini-cli-docs` provides official Gemini CLI documentation.

## Overview

This skill provides patterns for keeping Claude Code (`CLAUDE.md`) and Gemini CLI (`GEMINI.md`) memory files synchronized. The core principle is **CLAUDE.md as source of truth** with GEMINI.md importing and adding overrides.

## When to Use This Skill

**Keywords:** sync memory, sync context, claude.md gemini.md, memory import, context drift, @import, memport

**Use this skill when:**

- Setting up GEMINI.md for a new project
- Detecting drift between memory files
- Understanding the import syntax
- Troubleshooting sync issues

## Core Principle: Single Source of Truth

```text
CLAUDE.md (Source of Truth)
    │
    │ @import
    ▼
GEMINI.md (Imports + Overrides)
```

**Why CLAUDE.md is the source:**

- Claude Code is the primary development environment
- CLAUDE.md is already established in most projects
- Single point of update reduces maintenance burden
- Git history shows context evolution in one place

## GEMINI.md Structure

### Recommended Template

```markdown
# GEMINI.md

@CLAUDE.md

## Gemini-Specific Overrides

You are Gemini CLI. Your unique capabilities:
- Large context window (Flash) / Very large (Pro)
- Interactive PTY shell (vim, git rebase -i, htop)
- Checkpointing with instant rollback
- Policy engine for tool control
- Native Google Cloud authentication

### When to Use Your Strengths

- **Bulk analysis**: Use your large context for codebase-wide exploration
- **Interactive tools**: Handle vim, git interactive commands
- **Risky operations**: Use sandbox and checkpointing
- **Second opinions**: Provide independent validation

### Model Selection

- Use **Flash** for bulk analysis and simple tasks
- Use **Pro** for complex reasoning and very large contexts
```

### Import