| name | save-system |
| description | Design or implement save/load behavior, checkpoints, migrations, and corruption recovery. Use whenever state persistence changes. |
Save System
When to use
Design or implement save/load behavior, checkpoints, migrations, and corruption recovery. Use whenever state persistence changes.
Inputs to gather
- what must persist
- platform constraints
- migration risk
- rollback needs
Recommended roles
save_system_engineer
qa_lead
security_engineer
Primary docs / outputs
studio/docs/templates/save-migration-plan.md
studio/docs/templates/test-plan.md
Workflow
- Inspect the current repo/docs state first and cite concrete evidence.
- Choose the smallest useful output that moves the project forward.
- If code changes are involved, keep the patch tight and validate the changed behavior.
- Update or create durable docs when the result should persist.
- Recommend the next best role or skill if more work remains.
Category rules
- Map the actual files/systems first, then choose the smallest viable implementation slice.
- Prefer reversible, reviewable patches over broad speculative refactors.
- Report commands, tests, and manual checks honestly.
Deliverables
- save model
- migration notes
- corruption handling
- test matrix
Validation bar
- touched files/systems identified
- tests or manual validation listed
- follow-up work clearly scoped