Manage Context, Persistence, and Write-back

[Dumken1]

Main Question

How do we make local memory trustworthy, editable, persistent, and ready for write-back?

Fellow 4 Role

Fellow 4 owns control and trust.
This includes update, delete, persistence, and optional write-back after inference.

Branch

feature/m01w02-manage-context

Files Owned

src/python/local_first_ai/storage/manage_context.py
assets/screenshots/month-01-week-02/fellow-4-manage-context.png

Implementation Specification

Build APIs that manage context.
update_context_item(item_id, title=None, content=None, tags=None, importance=None)
delete_context_item(item_id)
prove_persistence()
write_back_context(context_type, title, content, source="inference_output")

The write-back function should simulate what happens after inference.

Example:
The model answers a user question.
The application extracts a useful summary.
The summary is saved as conversation_context or learning_record.

What Fellow 4 Must Prove

Fellow 4 should prove that:

1. A context item can be updated.
2. A context item can be deleted.
3. Invalid update ID does not crash the system.
4. Invalid delete ID does not crash the system.
5. Data remains after closing and reopening the database.
6. A simulated inference output can be written back as new context.