Generate an Improvement Plan
Generate a prioritised improvement plan from assessment gaps and work through it item by item, accepting, skipping, or deferring each action.
1. Run an assessment first (if you haven’t already)
The improvement plan works best when it has an existing assessment to draw gaps from. If you haven’t assessed the repository yet:
/assess
The assessment identifies your current level (L0–L5) and lists specific gaps. The improvement skill uses these gaps to filter the plan to only what is actually missing.
If you already know your level, you can skip the assessment and invoke the improvement skill directly.
2. Start the improvement plan
The improvement skill is invoked automatically at the end of /assess. To run it standalone:
/literacy-improvements
If invoked standalone, the skill asks for your current level:
What is your current AI literacy level?
1. Level 0 — Awareness
2. Level 1 — Prompting
3. Level 2 — Verification
4. Level 3 — Habitat Engineering
5. Level 4 — Specification Architecture
3. Choose a target level
The skill presents the levels above your current one:
You're currently at Level 2 (Verification).
How far would you like to improve?
1. Level 3 — Habitat Engineering (recommended next step)
2. Level 4 — Specification Architecture
3. Level 5 — Platform Engineering
The default recommendation is always the next level. Choosing a higher target includes all intermediate levels — the plan works through each level transition in order.
4. Work through the plan item by item
For each gap, the skill presents one item at a time with its priority and recommended action:
Improvement 1/6 (Level 2 → Level 3):
Gap: No HARNESS.md
Action: Run /harness-init
Priority: High — foundational; nothing else at L3 works without it
Accept / Skip / Defer?
Respond to each item:
- Accept — the command or skill runs immediately. Wait for it to complete before the next item appears.
- Skip — removes the item from the plan for this session.
- Defer — keeps the item in the plan but does not execute it now. The next assessment picks it up.
If an accepted command opens its own interactive flow (for example, /harness-init asks which features to enable), let it run naturally. The improvement plan resumes after the command completes.
5. Understand priority levels
| Priority | Criteria | Examples |
|---|---|---|
| High | Foundational — other items depend on it | HARNESS.md for L3, CI pipeline for L2 |
| Medium | Closes a real gap independently | Secret scanning, GC rules |
| Low | Conditional on project context | Docker scanning when no Docker is used |
Work through High items first. Medium and Low items can be deferred if resources are constrained.
6. Find the output document
After working through all items, the skill appends an Improvement Plan section to today’s assessment document if one exists:
assessments/YYYY-MM-DD-assessment.md
If no assessment exists for today, it writes a standalone file:
assessments/YYYY-MM-DD-improvements.md
The document records which improvements were accepted, skipped, or deferred, and what commands ran.
7. Re-run after completing deferred items
Deferred items are not lost — they appear again at the next assessment. When you have capacity to address them, re-run /assess or invoke /literacy-improvements directly to resume the plan.
Summary
After completing these steps you have:
- A prioritised list of improvements mapped to your specific gaps
- Commands executed for accepted improvements
- A record of accepted, skipped, and deferred items in the assessment document
- A clear target level and a path to reach it