New commands: - /summarize: Save session decisions/preferences to memory - /maintain: Configuration maintenance (validate, backup, restore, etc.) New automation scripts: - backup.sh: Create timestamped config backup (excludes secrets) - restore.sh: Restore config from backup (with validation) Updates: - component-registry.json: Added new commands - commands/README.md: Updated with new entries - automation/README.md: Documented utility scripts 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
102 lines
2.3 KiB
Markdown
102 lines
2.3 KiB
Markdown
---
|
|
name: summarize
|
|
description: Summarize and save the current session context
|
|
aliases: [save-session, session-summary]
|
|
---
|
|
|
|
# /summarize Command
|
|
|
|
Summarize the current session and save key information to memory.
|
|
|
|
## When Invoked
|
|
|
|
Extract and save important information from the current conversation:
|
|
|
|
1. **Decisions made** → `memory/decisions.json`
|
|
2. **Preferences learned** → `memory/preferences.json`
|
|
3. **Project context** → `memory/projects.json`
|
|
4. **Facts discovered** → `memory/facts.json`
|
|
|
|
## Process
|
|
|
|
### Step 1: Analyze Session
|
|
|
|
Review the conversation for:
|
|
- Explicit decisions ("let's use X", "we decided Y")
|
|
- Implicit preferences ("I prefer", "always do X")
|
|
- Project-related context (paths, configs, conventions)
|
|
- Facts about the environment or user
|
|
|
|
### Step 2: Categorize Items
|
|
|
|
For each item, determine:
|
|
- **Category**: decision, preference, project, fact
|
|
- **Content**: Brief description
|
|
- **Context**: Why it matters (optional)
|
|
|
|
### Step 3: Save to Memory
|
|
|
|
Append to the appropriate memory file:
|
|
|
|
```json
|
|
{
|
|
"id": "uuid",
|
|
"date": "YYYY-MM-DD",
|
|
"content": "Description of item",
|
|
"context": "Additional context",
|
|
"session": "session-id"
|
|
}
|
|
```
|
|
|
|
### Step 4: Update History Index
|
|
|
|
Mark session as summarized in `history/index.json`:
|
|
|
|
```json
|
|
{
|
|
"sessions": [
|
|
{
|
|
"id": "session-id",
|
|
"started": "timestamp",
|
|
"summarized": true
|
|
}
|
|
]
|
|
}
|
|
```
|
|
|
|
## Output Format
|
|
|
|
```
|
|
📝 Session Summary
|
|
|
|
Saved to memory:
|
|
• Decision: Use plugin structure for .claude config
|
|
• Preference: Prefer scripts over inline code in skills
|
|
• Fact: Gmail venv at ~/.claude/mcp/gmail/venv
|
|
|
|
Session marked as summarized.
|
|
```
|
|
|
|
## Memory File Locations
|
|
|
|
| File | Content |
|
|
|------|---------|
|
|
| `~/.claude/state/personal-assistant/memory/decisions.json` | Choices made |
|
|
| `~/.claude/state/personal-assistant/memory/preferences.json` | User preferences |
|
|
| `~/.claude/state/personal-assistant/memory/projects.json` | Project context |
|
|
| `~/.claude/state/personal-assistant/memory/facts.json` | Environment facts |
|
|
|
|
## When to Use
|
|
|
|
- End of a productive session
|
|
- After making important decisions
|
|
- When asked to "remember this"
|
|
- Before long breaks
|
|
|
|
## Policy
|
|
|
|
- Only save genuinely useful information
|
|
- Avoid redundancy with existing memory
|
|
- Keep descriptions concise
|
|
- Include context when helpful
|