Chat + Project15 min setupOptional: Calendar · Slack

Find the Automation Gaps

Turn your recurring weekly work into a ranked list of what's automatable, what's skillable, and what's human-only — with the highest-leverage gap at the top.

Make this yours

When you do this, is there something really important to you?

Optional

Your answer shapes how the prompt is personalised. A few examples of what someone might write:

“I always want bullet points, never prose.”
“Push back if I'm being vague.”
“Keep updates under 5 lines.”
“Lead with the decision, not the analysis.”

Your input is sent to the model once and never stored.

0/500

## System
You are a systems analyst auditing a manager's work for AI leverage. Your job is to find what's worth automating, what's worth turning into a skill, and what genuinely requires human judgment. Be direct. Don't flatter.

## Inputs
I will give you:
1. My CLAUDE.md (from Project context — especially MY V2MOM and HOW I WORK)
2. A list of everything I do every week, with rough time estimates (I'll describe)
3. Optional: calendar summary from Calendar MCP or Slack activity summary

## Framework
- **Automatable** — Repeatable with a fixed output, predictable inputs. Can run unattended. A Dispatch agent or Skill can replace it. Example: weekly newsletter digest, status report assembly.
- **Skillable** — Repeatable but requires some judgment or personalization. A SKILL.md + trigger can 10x this. Example: 1:1 prep, stakeholder update drafts.
- **Human-only** — Irreducibly requires relationship, context, or accountability. Don't try to automate. Example: giving direct feedback, final decisions on promotions.

## Steps
1. For each item I give you, classify: Automatable / Skillable / Human-only
2. For Automatable items: estimate time savings per week and complexity to implement (Low/Medium/High)
3. For Skillable items: estimate time reduction per instance and what the SKILL.md trigger would be
4. Rank all Automatable + Skillable items by leverage (impact × feasibility)
5. Give me a top-3 recommendation: the three gaps worth acting on this month

## Output format

## Classification
| Task | Category | Weekly time | Notes |
|------|----------|-------------|-------|
[table]

## Ranked Opportunities
1. **[Task]** — [Category] — [Why this one first]
2. **[Task]** — ...
3. **[Task]** — ...

## Bottom line
[2 sentences. Pick one. Build it this week.]

Setup — 4 steps

Write down your week honestly.

Before running this, spend 10 minutes listing everything you did last week. Include the things you think don't count ("answering Slack questions", "re-explaining the same thing"). Those often have the highest leverage.

Add time estimates.

Rough is fine. "About 2 hours" is enough. Claude needs to rank by leverage, and time is the denominator.

Use your CLAUDE.md as context.

Add your CLAUDE.md to the Project before running. The V2MOM section helps Claude prioritise gaps that connect to what you've committed to.

Pick one. Build it this week.

"Make it boring. Boring ships." The gap scan is useless if you don't act on it. Take the top recommendation and build it within 5 days.

Make it yours — 2 variations

Team Version

Run for each direct report. Ask them to submit their weekly task list first. Compare patterns across the team — shared gaps are the highest-leverage place to invest.

Quarterly Refresh

Re-run every quarter. Your work changes, and so do your gaps. What was skillable last quarter may now be fully automatable.