---

name: update description: Save what you worked on so your AI can pick it up next time

Session Update

Run this when you're finishing a work session, or at natural breakpoints during longer sessions.

Invocation: /update or "Let's save our progress"

Instructions

When this skill is invoked, follow these steps:

1. Get Current Date and Time

Use the system date to determine:

• Today's date (for the log filename): YYYY-MM-DD
• Current time (for the session entry): HH:MM (24-hour format)

2. Create or Update Today's Session Log

Session logs live in ~/ai-data/logs/sessions/YYYY-MM-DD.md.

If the file doesn't exist, create it with this structure:

Copy# Session Log - YYYY-MM-DD

## Summary
**Major Accomplishments**:
- (To be filled)

## Next Actions
- (To be filled)

---

## Sessions

### HH:MM - [Brief description]
- What we worked on
- Key decisions made
- Files created or modified
- Open questions or next steps

If the file exists, add a new session entry after the --- separator, before previous sessions (reverse chronological order).

3. Capture the Session Content

Document:

• What was worked on - tasks, problems, topics discussed
• Key decisions - choices made and why
• Concrete outcomes - files changed, code written, problems solved
• Open items - things to pick up next time

Be specific but concise. Future-you needs enough detail to remember context.

4. Update the Summary Section

Add or update the summary at the top of the file to reflect today's accomplishments.

5. Update Next Actions (if applicable)

If there are clear next steps, add them to the Next Actions section.

6. Capture Observations (Optional)

If you noticed something about how the user prefers to work:

• Add it to ~/ai-data/learning/user_model.yaml
• Use the format from ~/ai-data/learning/templates/user_model_template.yaml

Only capture genuine observations, not every interaction.

7. Confirm Completion

Brief acknowledgment: "Session logged." or "Saved to today's log."

---

Example Session Entry

Copy### 14:30 - Debugging API timeout issue
- Investigated timeout errors in production logs
- Root cause: Connection pool exhaustion under load
- Increased pool size from 10 to 50 in config/database.yml
- Deployed to staging, monitoring for 30 minutes
- TODO: Deploy to production if staging looks good

---

Customization Ideas

As you get comfortable, you might add:

• Reflection prompts ("What worked well?")
• Automatic git commits of the session log
• Cross-posting to daily notes or task systems

These are enhancements - the core skill works without them.